3 pursley 801 - 865 as of 6-4 63... · 3 pursley 801 - 865 as of 6-4 (do not delete) 6/4/2012 3:36...

3 PURSLEY 801 - 865 AS OF 6-4 (DO NOT DELETE) 6/4/2012 3:36 PM

DEFEASIBLE FEDERALISM

Garrick B. Pursley∗

ABSTRACT

What does it mean for federalism—obviously an issue of constitutional magnitude—to influence constitutional doctrine without being the object of constitutional doctrine? This phenomenon is familiar—concern about the effects of government action or constitutional doctrine on the viability of the federalist system influence doctrine that implements the Commerce Clause, the Supremacy Clause, constitutional rights provisions, and others. But constitutional theory lacks a thorough account of the conceptual status of what courts call, variously, federalism “impacts,” “effects,” “concerns,” and so forth in these instances. Federalism does not always have the decisive weight conventionally attributed to constitutional norms—it functions as a defeasible reason for decision in some situations. How can a constitutional norm be overridable by non-constitutional—even non-legal—considerations? I develop a new account on which the strength of federalism norms as reasons for decision varies from decisive where the stakes for the stability of the constitutional structure are high to overridable where the structural stakes are lower but the public policy stakes are high. This view of federalism norms departs from conventional accounts but, I argue, better explains the different roles we observe federalism norms playing in adjudication. ABSTRACT .................................................................................................. 801 INTRODUCTION .......................................................................................... 802 I. FEDERALISM’S ROLES ..................................................................... 808

A. Direct Federalism Rules .......................................................... 813 B. Indirect Federalism Rules ....................................................... 819 C. Hard Cases .............................................................................. 830

∗ Assistant Professor, The University of Toledo College of Law (Fall 2010–Summer 2012); Assistant Professor, The Florida State University College of Law (beginning Fall 2012). I am grateful to Mitch Berman, Allan Erbsen, Ian Farrell, Willy Forbath, Sandy Levinson, Ernie Young, and Hannah Wiseman for helpful discussions about versions of this piece; and I must thank Mark Brandon, Jelani Jefferson-Exum, Jessica Knouse, Dan Markel, Rob Mikos, Bill Richman, Jim Rossi, Mark Seidenfeld, Lee Strang, Franita Tolson, Rebecca Zietlow, and the participants in workshops at Vanderbilt University Law School and the University of Toledo College of Law, for helpful comments and suggestions on later versions. This is dedicated to Hannah Pursley, who was born shortly after I started thinking about these issues and whose joyful influence has changed my thinking on nearly everything.


802 Alabama Law Review [Vol. 63:4:801

II. FEDERALISM’S DEFEASIBILITY ....................................................... 835 A. Against the Standard Model .................................................... 837 B. Instrumental Reasons and Underenforcement ......................... 846 C. Federalism as a “Quasi-Constitutional” Reason .................... 854

1. Intermediate Normative Priority ....................................... 854 2. Variable Normative Priority .............................................. 860

CONCLUSION .............................................................................................. 866

INTRODUCTION

What does it mean for federalism—unquestionably a constitutional consideration—to influence constitutional doctrine in cases where it is not the object of constitutional doctrine? My descriptive claim that this occurs is not controversial. Concerns about the effects of government action or judge-crafted constitutional doctrine on the stability of the constitutional system have influenced doctrinal formulation and outcomes in Commerce Clause cases, dormant Commerce Clause cases, preemption cases, constitutional rights cases, and other areas in which courts do not directly enforce federalism norms.1 In these instances, federalism seems to operate not as a decisive constitutional reason for decision, but instead as one in a category of sub-constitutional considerations—including things like institutional capacity, interbranch relations, and other pragmatic considerations—bearing on the doctrinal rule, test or standard the court formulates to implement the non-federalism norm. My more ambitious claim—the “variable normativity hypothesis”—is that we may better understand and explain the role of federalism norms in these instances by viewing constitutional federalism norms as defeasible, rather than determinative, reasons for decision under certain circumstances. Put differently, the normative force of federalism norms appears to shift between decisive, like other constitutional norms, and overridable, like pragmatic considerations courts consider in constitutional cases, depending on the circumstances. This cuts against the conventional characterization of constitutional norms as superordinate, decisive legal reasons for decision where they apply; but I will argue that it explains the phenomena better than possible alternative accounts. The variable normativity hypothesis is distinct from conventional “functionalism”—roughly, the view that

1. Cf. Ernest A. Young, The Rehnquist Court’s Two Federalisms, 83 TEX. L. REV. 1, 8 (2004) (characterizing as federalism doctrine several rules that only indirectly rely on federalism, including the “‘dormant’ commerce principle,” “the abstention doctrines, rules of federal common lawmaking, the adequate and independent state grounds bar . . . rules governing federal habeas corpus review of state convictions and virtually the whole corpus of conflict of laws” rules) [hereinafter Young, Two Federalisms].


2012] Defeasible Federalism 803

constitutional adjudication does or should proceed by giving weight to current practicalities in the interpretation and application of constitutional norms—because it suggests that in a number of instances federalism does, in fact, carry non-defeasible normative force. However, if accurate, it does justify many functionalist proposals.

The variable normativity hypothesis advances recent work on the conceptual structure of constitutional adjudication by suggesting that an additional category of “quasi-constitutional” reasons legitimately bears on the formulation of constitutional doctrine. And by providing new analytic tools for understanding judicial engagement with federalism in unusual circumstances, it also provides new leverage on central debates in federalism scholarship and constitutional theory.

Federalism is an ineliminable feature of the American constitutional structure.2 On that observation, we might reasonably expect to find unflinching norms that subtend the federalist structure, perhaps by marking off some mandatory allocation of federal and state authority, some non-transferable federal and state government functions, etc. But in examining judicial engagement with federalism, we find seemingly defeasible constitutional norms. The presumption against preemption, for example, is a rule of statutory interpretation that courts periodically use to determine the preemptive scope of federal law. It instructs courts not to construe statutes to preempt state law absent clear evidence of congressional intent.3 Federalism-related concerns about the constricting effects of preemption on state regulatory authority partially justify the presumption, but federalism norms are not its direct object. The presumption is an interpretive canon; it is one of a category of rules whose primary purpose is to determine statutory meaning.4 The odd thing here is that the federalism concerns that partially shape the rule, though significant, are overridable by clear legislative language or unambiguous congressional intent; and the presumption is not applied at all in some preemption cases.5 The

2. See GEOFFREY R. STONE ET AL., CONSTITUTIONAL LAW 167–69 (6th ed. 2009) (describing federalism as a clear structural presupposition of the Constitution); New York v. United States, 505 U.S. 144, 157 (1992) (explaining that federalism is part of “the framework set forth in the Constitution”); U.S. Term Limits, Inc. v. Thornton, 514 U.S. 779, 838 (1995) (Kennedy, J., concurring) (noting that the Constitution established for Americans “two political capacities, one state and one federal”).

3. Rice v. Santa Fe Elevator Corp., 331 U.S. 218, 230 (1947) (“[T]he historic police powers of the States [are] not to be superseded . . . unless that was the clear and manifest purpose of Congress.”); see Stephen A. Gardbaum, The Nature of Preemption, 79 CORNELL L. REV. 767 (1994) (discussing the Rice presumption).

4. See generally William N. Eskridge, Jr. & Philip P. Frickey, Quasi-Constitutional Law: Clear Statement Rules as Constitutional Lawmaking, 45 VAND. L. REV. 593 (1992).

5. Federalism norms might require a presumption against preemption and no more. This might flow from the controversial “political safeguards” view of federalism. See infra notes 146–150 and accompanying text. Failure to apply the presumption in all preemption cases suggests that even this



presumption cannot directly enforce any mandatory allocation of authority between national and state governments—non-constitutional considerations override federalism to change that allocation in significant ways.6 Similar examples appear in other contexts.

The federalism literature is rich with normative claims about how federalism should figure in adjudication. But basic questions remain about how federalism norms actually do function in various adjudicatory settings. Federalism is a proper object of constitutional doctrine—few doubt that some constitutional federalism norms exist;7 and several doctrines straightforwardly enforce those norms by invalidating violations.8 These “direct federalism rules” map easily onto the standard model of constitutional adjudication, in which courts identify and invalidate violations of constitutional norms.9 The standard model suggests that nothing can outweigh an identified and cognizable constitutional violation in judicial decisionmaking—constitutional norms are decisive such that identified violations always require invalidation.10 The presumption against preemption and similar doctrines, however, show that federalism norms sometimes influence the formulation of constitutional doctrine in a manner distinct from the conventional “invalidate identifiable violations”

version of the norm is treated as overridable. See generally Mary J. Davis, The “New” Presumption Against Preemption, 61 HASTINGS L.J. 1217, 1219 (2010).

6. Cf. Mitchell N. Berman, Constitutional Decision Rules, 90 VA. L. REV. 1 (2004) (highlighting the role of non-constitutional considerations in the formulation of constitutional doctrine) [hereinafter Berman, Rules].

7. Strict textualists might reject the legitimacy of inferring any structural norms. See, e.g., John F. Manning, Federalism and the Generality Problem in Constitutional Interpretation, 122 HARV. L. REV. 2003 (2009). But this critique of federalism doctrine is controversial; few credit the strict textualist method. See, e.g., Gillian Metzger, The Constitutional Legitimacy of Freestanding Federalism, 122 HARV. L. REV. F. 98 (2009). There is significant interpretive debate about the content of federalism norms, but my point is that most agree that federalism norms exist.

8. See, e.g., New York v. United States, 505 U.S. 144, 157 (1992) (holding that federalism norms must be enforced “even if one could prove that federalism secured no advantages to anyone”). See generally Ernest A. Young, Making Federalism Doctrine: Fidelity, Institutional Competence, and Compensating Adjustments, 46 WM. & MARY L. REV. 1733 (2004) (defending rules for making federalism doctrine) [hereinafter Young, Federalism Doctrine].

9. See Berman, Rules, supra note 6, at 4, 5–6, 32–33 (2004) (discussing the standard model); Garrick B. Pursley, The Structure of Preemption Decisions, 85 NEB. L. REV. 912, 919–20 (2007) [hereinafter Pursley, Preemption Decisions] (same).

10. See Marbury v. Madison, 5 U.S. (1 Cranch) 137, 177–80 (1803) (announcing that constitutional provisions are “superior, . . . paramount law” in relation to other forms of federal and state law). Some constitutional norms are not judicially enforceable, see, e.g., Nixon v. United States, 506 U.S. 224, 226 (1993) (Impeachment Clauses); Luther v. Borden, 48 U.S. 1, 42 (1849) (Guarantee Clause), others are underenforced by doctrines that invalidate fewer actions than actually violate the underlying norm, see Lawrence Gene Sager, Fair Measure: The Legal Status of Underenforced Constitutional Norms, 91 HARV. L. REV. 1212 (1978). These indirect federalism rules appear to present a distinct phenomenon, however—namely, one in which doctrine is not designed to identify all, or a subset of, violations of federalism norms but rather to account for federalism in some different sense in the process of identifying violations of non-federalism norms.



formula—under certain circumstances, federalism is a defeasible influence on doctrinal rules.

Indirect federalism rules are familiar. We frequently say that federalism “values” or “principles” effect outcomes in all sorts of cases.11 Arguments about the constitutional structure, including federalism arguments, have been characterized as a principal “modality” of constitutional reasoning relevant in many cases.12 There is a normative literature concerned with the desirability of structure’s wide-ranging influence on doctrinal development in non-structural cases.13 We lack a thorough conceptual account of what we refer to when we say that a court invoked federalism “values” or “principles” in these indirect contexts—we seem to refer to constitutional norms exerting defeasible influence on doctrine. But how can constitutional norms can function this way? These “indirect federalism rules” are the phenomena I want to examine. The indirect federalism rules I discuss below support a “Defeasible Federalism Thesis” (DF): “Constitutional federalism norms assert defeasible rather than decisive influence on the formulation of constitutional doctrine and the outcome of constitutional adjudication under certain circumstances.” My goal is to determine whether there is a theoretical justification for the proposition, compelled by observing judicial decision-making, that DF is true.

I make three contributions: First, by sorting a broad cross-section of federalism-influenced doctrine into the direct and indirect categories, I carve the terrain in a new way that usefully highlights this under-explained phenomenon. Second, I assess several theoretical predicates and the explanations for indirect federalism rules that they might support. Third, I develop a new explanation for indirect federalism rules based on the variable normativity hypothesis. My solution is to depart from the standard model of constitutional adjudication only insofar as that model suggests

11. See, e.g., STONE ET AL., supra note 2, at 167–69 (discussing federalism “values”); Young, Two Federalisms, supra note 1, at 8 (same). A parallel observation holds for separation-of-powers norms, which often effect doctrinal formulation even when they are not its direct object. See, e.g., Ashwander v. Tenn. Valley Auth., 297 U.S. 288, 346–48 (1936) (Brandeis, J., concurring) (explaining interpretive canon requiring courts to avoid constitutional decisions where possible); Crowell v. Benson, 285 U.S. 22, 62, 62 n.30 (1932) (similar); Cass R. Sunstein, Interpreting Statutes in the Regulatory State, 103 HARV. L. REV. 405, 468–69 (1989); Ernest A. Young, Constitutional Avoidance, Resistance Norms, and the Preservation of Judicial Review, 78 TEX. L. REV. 1549, 1573 (1999) [hereinafter Young, Resistance Norms] (both noting that avoidance canons are partially justified by separation-of-powers concerns).

12. See generally PHILIP BOBBITT, CONSTITUTIONAL FATE: THEORY OF THE CONSTITUTION 74–92 (1982) (arguing that “we are held captive by the idea that a specific text or doctrine engrafted onto text is the only sure guide to constitutional decisionmaking, even when structural argument is doing the real work of resolving the issue”).

13. Compare, e.g., Young, Federalism Doctrine, supra note 8 (arguing that federalism should weigh in a variety of adjudicatory contexts), with Eskridge & Frickey, supra note 4, at 629–45 (criticizing this approach).



constitutional norms must always carry superordinate normative status. Federalism norms, I argue, appear defeasible in indirect cases because they are defeasible in those cases: They may have differing degrees of normative force, and thus function as something other than standard constitutional norms, under certain circumstances.

Constitutional theorists have begun considering norms that do not arise directly from the canonical Constitution yet function in some sense as constitutional norms. For example, certain “super statutes” that perform constitutive functions like structuring government agencies or protecting rights are treated as superordinate against ordinary law.14 This suggests a category of legal norms possessed of normative force somewhere between that of standard normatively superordinate constitutional norms and ordinary legal norms. I argue that this category of intermediate normative force may include norms, like federalism norms, that do arise directly from the canonical Constitution. Courts sometimes properly treat federalism norms as something other than “higher” law and, when they do, they may legitimately weigh non-constitutional and even non-legal factors more heavily than considerations of federalism. The important analytic task that remains after constructing an explanation predicated on the variable normativity hypothesis is to isolate the factors that determine whether federalism functions defeasibly. Preliminarily, I argue that federalism tends toward decisive normative force where a challenged action presents a significant risk to the basic constitutional structure and toward defeasible where the structural risk is minor and the non-federalism constitutional or substantive policy implications of the decision are significant.

This account fits the phenomena better, is simpler, and is more consistent with other well-founded beliefs about constitutionalism than alternatives. It new nuance to our framework for understanding and constructing federalism doctrine and recasts some persistent normative debates. The variable normativity account might, for example, provide a new justification for approaches to federalism I have called “compatibilist”; a cluster of positions loosely bound by the claim that federalism doctrine should for pragmatic reasons accommodate innovative intergovernmental arrangements in environmental protection, energy conservation, labor, immigration, and other areas.15 Compatibilism is challenged by claims that

14. See generally William N. Eskridge, Jr. & John Ferejohn, Super-Statutes, 50 DUKE L.J. 1215, 1216 (2001) (explaining this concept).

15. Garrick B. Pursley, Federalism Compatibilists, 89 TEX. L. REV. 1365, 1367–69 & nn.11–15, 20–23 (2011) (reviewing ROBERT A. SCHAPIRO, POLYPHONIC FEDERALISM: TOWARD THE PROTECTION

OF FUNDAMENTAL RIGHTS (2009)) (canvassing compatibilist literature); see, e.g., SCHAPIRO, supra (describing and defending a system of “polyphonic” federalism involving federal-state cooperation); Jessica Bulman-Pozen & Heather K. Gerken, Uncooperative Federalism, 118 YALE L.J. 1256, 1259 (2009) (arguing that states wield “the power of the servant” rather than autonomy-based checks on



federalism norms unavoidably require structures inconsistent with flexibility. The variable normativity hypothesis, however, highlights contexts in which structural norms are amenable to pragmatically motivated modification. Indeed, viewing federalism norms as functionally similar to instrumental influences in some doctrinal settings broadens federalism’s capacity; it suggests that federalism is relevant in more constitutional cases than we thought.16

This new account of federalism norms also adds to theoretical work on the distinction between constitutional interpretation and the formulation of the rules, tests, and standards that implement constitutional norms.17 This literature documents and defends the influence of non-constitutional considerations—including, for example, concerns about institutional competence and adjudicatory error—on these implementing rules.18 Federalism considerations are conceptually distinct from these standard instrumental influences on doctrine—they are, after all, grounded on constitutional norms; an account of the ways in which federalism norms legitimately influence non-federalism doctrine adds a new category of “quasi-constitutional reasons” to our model of doctrinal formulation. Finally, the variable normativity explanation helps to simplify structural theory in general. It justifies most federalism-influenced doctrines without positing multiple complex federalism norms. A single, simple norm—e.g., “there must be rough balance of authority between the national and state governments”19—that varies in normative force, combined with instrumental considerations, can explain nearly every federalism-related doctrine that we observe. These doctrinal rules may be shaped by a complex amalgam of non-constitutional reasons, but their normative predicate may be a simple proposition on which nearly everyone can agree.20 This way of explaining the formulation of constitutional law is an

federal power); Brian Galle & Mark Seidenfeld, Administrative Law’s Federalism: Preemption, Delegation, and Agencies at the Edge of Federal Power, 57 DUKE L.J. 1933 (2008) (arguing for a “constitutional realism” accommodating flexible arrangements); ERWIN CHEMERINSKY, ENHANCING

GOVERNMENT: FEDERALISM FOR THE 21ST CENTURY (2008) (defending a system of “empowerment federalism”).

16. Where federalism norms function as defeasible reasons for doctrinal formulation, their weights relative to other non-constitutional reasons varies as well. Metadoctrinal theory lacks firm criteria for weighting various instrumental determinants of doctrine, but adding quasi-constitutional reasons suggests a rough ordering. Presumptively courts should weigh interpretive and quasi-constitutional reasons more heavily than standard instrumental considerations, but this presumption should be overridable.

17. See generally sources cited infra note 28. 18. See Berman, Rules, supra note 6, at 50, 57–60. On this distinction and its uses and

implications, see generally sources cited infra note 28. 19. See Young, Federalism Doctrine, supra note 8 (proposing this as a hypothetical norm). 20. See Garrick B. Pursley, Dormancy, 100 GEO. L.J. 497 (2012); see also Pursley, Federalism

Compatibilists, supra note 15.



advance—indeed, if non-constitutional legal norms are legitimate in virtue of public and official consensus on their validity and content, then pluralistic societies may not be able to manage more than this kind of thin structural norm.

I don’t mean to suggest that there is a single explanation for all indirect federalism rules. Some may be instances of judicial underenforcement of federalism norms or other practices acknowledged by conventional accounts; but no conventional account captures all the phenomena. In Part I, I discuss examples of direct and indirect federalism rules to clarify the distinction. Part II explores potential explanations for indirect federalism rules. I first assess several explanations based on several established positions in constitutional theory, including varieties of constitutional realism, process models of adjudication, and the theory of underenforced constitutional norms. Identifying problems with these explanations, I develop the variable normativity hypothesis as the conceptual predicate for an alternative explanatory account and canvas its benefits—both for the project of understanding doctrine and for central debates in the federalism and constitutional theory literatures. A brief Conclusion follows.

I. FEDERALISM’S ROLES

Federalism is under-specified in the canonical Constitution. Its metes and bounds must be inferred; thus doctrine designed to implement federalism norms is created at some analytical distance from federalism’s textual foundations.21 Changing circumstances require modification of doctrines that structure and sustain the government.22 As one might expect in these circumstances, federalism doctrine is multifaceted and difficult to get one’s arms around.23

There are various ways to carve it up.24 Mine is to distinguish “direct” from “indirect” federalism doctrines. Direct federalism doctrines purport to

21. See Young, Federalism Doctrine, supra note 8, at 1775–83 (discussing federalism doctrine’s detachment from the spare constitutional text); Pursley, Dormancy, supra note 20, at 514–19 (discussing structural inferences).

22. See Pursley, Dormancy, supra note 20, at 512–14 (arguing that a basic goal of the constitutional system is to achieve long-term durability); Young, Federalism Doctrine, supra note 8, at 1750–58 (arguing that the federal structure has been and should be adjusted in response to changing circumstances).

23. E.g., Matthew D. Adler & Seth F. Kreimer, The New Etiquette of Federalism: New York, Printz, and Yeskey, 1998 SUP. CT. REV. 71, 71 (arguing that federalism doctrine “lacks a fabric of constitutional law sufficiently coherent and well-justified to last”); Young, Two Federalisms, supra note 1, at 6 (“It is customary to start by saying that the Supreme Court has failed to develop a coherent theory of federalism.” (footnote omitted)).

24. Professor Young has a comprehensive taxonomy that divides federalism rules along three dimensions. See Young, Two Federalisms, supra note 1, at 13–18 (categorizing federalism doctrine according to “the aspect of federalism to be promoted, the focus of judicial review on issues of



directly enforce federalism norms by identifying and invalidating violations of those norms. Indirect federalism doctrines do not appear to directly enforce federalism norms but nevertheless appear to be influenced by federalism considerations. I flesh out the direct/indirect federalism doctrine distinction in this Part with examples from both the vertical and horizontal dimensions of federalism—the allocation of authority between the federal and state governments and the relationships between state governments, respectively.25 I conclude by examining cases that resist easy categorization and highlighting the reasons that indirect federalism rules need further explanation.

Commentators tend to characterize indirect federalism doctrines as influenced by “principles” of federalism or designed in part to promote “values” of federalism in addition to their primary purposes. But we lack an account of the conceptual status of these federalism “values” or “principles,” the way in which federalism norms influence these doctrines, and the implications of both the existence of constitutional “values” or “principles,” and of this indirect form of normative influence for constitutional theory. I argue that the best explanation is that these indirect doctrines are influenced by federalism norms—not a distinct constitutional input, a “value” or “principle.” Federalism norms influence constitutional doctrine without mandating certain formulations or outcomes in the decisive manner we conventionally attribute to constitutional norms. Federalism norms in these instances act as defeasible reasons in doctrine-making. The central conceptual question is how constitutional norms may act as defeasible reasons in the light of our standard model of constitutional adjudication.

The standard model has three steps: First, courts articulate, typically through a process of interpretation, the content of the applicable constitutional norm; second, they craft a mediating rule to implement that norm in concrete cases; and third, they apply the mediating rule to the facts to generate a constitutional holding.26 An important corollary is that constitutional norms are accorded superordinate normative status in the hierarchy of legal authorities; courts presume that constitutional norms trump any other relevant legal or non-legal considerations and thus exert

substance or process, and the rigidity of judicial review in terms of the ease with which other actors may override its results”).

25. See Allan Erbsen, Horizontal Federalism, 93 MINN. L. REV. 493, 494 (2008) (noting division of power along “a vertical plane that establishes a hierarchy and boundaries between federal and state authority, and a horizontal plane that attempts to coordinate fifty coequal states that must peaceably coexist.”).

26. See Berman, Rules, supra note 6, at 4, 5–6, 32–33 (2004) (describing the standard model).



decisive influence on the outcomes of cases in which they apply.27 A recent insight is that the second step—formulating mediating decision rules—involves a variety of considerations:28 Courts consider interpretive factors, most importantly the rules’ “fit” with the proper construction of the underlying constitutional norm; but they also consider instrumental factors such as rules’ workability, their potential to generate errors, comparative institutional capacity, interbranch friction, and so forth.29 These instrumental considerations are non-legal and thus overridable by legal requirements. But it is difficult to explain the contours of constitutional doctrine based solely on legal reasons. I argue that federalism and other structural constitutional requirements function more like the instrumental considerations than the interpretive considerations in some instances. Commentators working on doctrinal formulation have implicitly acknowledged the possibility of federalism norms’ functioning defeasibly in certain cases by including structural concerns among the category of instrumental influences on doctrine.30

27. See, e.g., Marbury, 5 U.S. (1 Cranch) at 177–180 (characterizing constitutional norms as “superior, paramount law” in relation to other forms of law); McCulloch v. Maryland, 17 U.S. (4 Wheat.) 316 (1819) (holding federal constitutional provisions superordinate over state law); see also Bradford N. Clark, The Supremacy Clause as a Limit on Federal Power, 71 GEO. WASH. L. REV. 91, 92–93 (arguing that the Constitution’s Supremacy Clause in Article VI establishes that constitutional requirements trump ordinary federal and state law); Carlos E. Gonzalez, The Logic of Legal Conflict: The Perplexing Combination of Formalism and Anti-Formalism in Adjudication of Conflicting Legal Norms, 80 OR. L. REV. 447, 479–81 (2001) (arguing that courts “always and unconditionally preference norms belonging to a legal category of superordinate position in the ordering of legal categories over norms belonging to a legal category of subordinate position in that ordering” and that constitutional norms occupy the highest rank in that ordering, such that “[w]here . . . a constitutional norm prohibits X and a statutory, administrative, or common law norm allows X, courts always and unconditionally privilege the constitutional norm over the statutory, administrative, or common law norm, and find that the law ultimately prohibits X”). This model of constitutional adjudication leaves out some nuance; but my point is only that it is widely accepted. For one thing, the question of the proper method for carrying out step 1—constitutional interpretation—is glossed over here; that question, of course, is at the center of controversy in modern constitutional theory.

28. A growing literature explores this process of doctrinal rule formulation. See generally Berman, supra note 6; Kermit Roosevelt III, Constitutional Calcification: How the Law Becomes what the Court Does, 91 VA. L. REV. 1649, 1658 (2005); Richard H. Fallon, Jr., The Supreme Court 1996 Term—Foreword: Implementing the Constitution, 111 HARV. L. REV. 54 (1997); Henry P. Monaghan, The Supreme Court, 1974 Term—Foreword: Constitutional Common Law, 89 HARV. L. REV. 1, 31 (1975); Sager, supra note 10, at 1212. Some commentators argue that all doctrine is equally influenced in its formulation by non-legal considerations. See Berman, supra note 6, at 43–51 (discussing these “pragmatists”). See generally Daryl J. Levinson, Rights Essentialism and Remedial Equilibration, 99 COLUM. L. REV. 857 (1999); David A. Strauss, The Ubiquity of Prophylactic Rules, 55 U. CHI. L. REV. 190 (1988). I have discussed this feature of the debate elsewhere and will not revisit it here, see Pursley, Dormancy, supra note 20, at 502–12. Regardless, we still may usefully distinguish among kinds of constitutional doctrine “even while conceding the legitimacy of each, and without staking ourselves to any claims about the sorts of considerations upon which courts might rely in the derivation and formulation of either.” Berman, supra, at 60.

29. Berman, supra note 6, at 30–36. 30. See, e.g., Sager, supra note 10, at 1219–22.



The concept of a constitutional “norm” is often under-explained. Conventionally, we take norms to be constitutional requirements, e.g., the bicameralism and presentment requirements for enacting federal legislation;31 prohibitions, e.g., the rule against deprivations without due process of law;32 permissions, e.g., the grant of congressional authority to regulate interstate commerce33 or the reservation of general police power to the states;34 etc. The conventional understanding of constitutional “norms” may be expressed as constitutional requirements, permissions, prohibitions, and so forth, simpliciter. Many constitutional requirements, prohibitions, and permissions must be adduced through a contested process of interpretation and thus interpreters may disagree about their content.35 But typically we do not debate constitutional norms’ status as superordinate over other considerations, including non-constitutional legal norms, where they apply—the well-accepted conventional view “that courts must make decisions according to law” suggests that courts must “privilege constitutional text and, to a lesser extent, history over other more functional or consequentialist sources of doctrine.”36

That is not to say that constitutional norms are always fully enforced. Courts in some instances enforce norms less than fully or overlook violations, reign in permissions, or even overenforce prohibitions.37 Judicial underenforcement of constitutional norms is fairly well accepted38 and provides a potential explanation for indirect federalism rules. But even when underenforced by courts, constitutional norms are binding to their conceptual limit on other actors and retain their full trumping status in and out of court.39 Underenforcement is justified not by intrinsic characteristics of constitutional norms, but by instrumental concerns about the functioning of courts in constitutional cases.40 In indirect cases, however, federalism

31. U.S. CONST., Art. I, § 7. 32. U.S. CONST., Amdts. V, XIV. 33. U.S. CONST., Art. I, § 8, cl. 3. 34. U.S. CONST., Amdt. X. 35. One might fairly say that disputes over interpretive method are the central disputes in modern

constitutional theory. For an overview, see generally Daniel A. Farber, The Originalism Debate: A Guide for the Perplexed, 49 OHIO ST. L.J. 1085 (1989); Mitchell N. Berman, Originalism Is Bunk, 84 N.Y.U. L. REV. 1 (2009).

36. Young, Federalism Doctrine, supra note 8, at 1837. 37. There are, of course, important questions about the legitimacy of these enforcement practices

that I set aside here. For an overview, see sources cited supra note 28. 38. See Sager, supra note 10. 39. See Sager, supra note 10, at 1221 (“[C]onstitutional norms which are underenforced by the

federal judiciary should be understood to be legally valid to their full conceptual limits, and federal judicial decisions which stop short of those limits should be understood as delineating only the boundaries of the federal courts’ role in enforcing the norm.”).

40. See Sager, supra note 10, at 1217–18 (discussing “analytical and instrumental” reasons for judicial underenforcement); see also infra Part II.B.1.



itself appears to act as an instrumental side-constraint modulating the court’s approach enforcing other norms. Thus, the possibility of underenforcement, without more, does not explain indirect federalism rules—at least not obviously. Even with underenforcement, the conventional model lacks some vocabulary or concepts needed to assign constitutional norms as something less than full superordinate normative status. Yet that is what we observe in the indirect federalism cases—federalism, there, functions as a weighty but defeasible influence that may be outweighed by other considerations. This is the puzzle.

This is not just a question about federalism. Other constitutional norms appear defeasible in some contexts—think here of basic separation-of-powers norms41 and fundamental systemic norms requiring, for example, minimal process prior to governmental coercion, minimal material and social equality, and representative government processes.42 Federalism’s visibility makes it a useful context in which to explore this potentially generalizable phenomenon. The tension with the standard model suggests that the conceptual question presented by indirect federalism rules is both open and interesting. To phrase it more precisely, let us generalize these observations as the “Defeasible Federalism Thesis” (DF): “Constitutional federalism norms assert defeasible rather than decisive influence on the formulation of constitutional doctrine and the outcome of constitutional adjudication under certain circumstances.”

I offer a new explanation for the apparent truth of DF that draws in part on recent developments in constitutional theory to argue that federalism

41. As with federalism, we often say that separation-of-powers principles or values influence the outcomes in cases directly concerned with other matters, e.g., in statutory interpretation cases, administrative law cases, or cases involving the scope of the appointments power. Cf. Victoria Nourse, Misunderstanding Congress: Statutory Interpretation, the Supermajoritarian Difficulty, and the Separation of Powers, 99 GEO. L.J. 1119, 1122 (2011) (“Theories of statutory interpretation [assume] . . . normative theories about how Congress should relate to courts or agencies. In short, theories of statutory interpretation assume, often without any justification or articulation, theories of the separation of powers.”); John F. Manning, Separation of Powers as Ordinary Interpretation, 124 HARV. L. REV. 1939, 1942 (2011) (noting that “every statutory scheme entails some choice about the distribution of power between or among branches, [thus] the composition of virtually any federal instrumentality potentially raises questions under the separation of powers doctrine”).

42. See, e.g., LAWRENCE G. SAGER, JUSTICE IN PLAINCLOTHES: A THEORY OF AMERICAN

CONSTITUTIONAL PRACTICE 145–46 (2004) (making a conceptual case for basic constitutional norms requiring equality of membership in the political community, fair and open government processes, for all “the opportunity to secure materially decent lives,” and for all “reasonable latitude in leading the lives they choose to lead”); SOTIRIOS A. BARBER, WELFARE AND THE CONSTITUTION 1–22 (2003) (arguing that the Constitution entrenches norms requiring government action to enhance citizens’ welfare); JOHN HART ELY, DEMOCRACY AND DISTRUST: A THEORY OF JUDICIAL REVIEW (1980) (arguing that constitutional review in courts should be centered on the goal of reinforcing a basic constitutional norm requiring robust democratic deliberation about legal and social questions). We need to be careful here. On the possibility that there exists some distinction between these kinds of guarantees as examples of constitutional “principles” rather than “norms,” see infra notes 208–213 and accompanying text.



plays a role in the formulation of indirect federalism rules distinct from that of other kinds of influences on constitutional doctrine—interpretive and instrumental reasons—because federalism is the kind of constitutional norm that may be legitimately treated as having something less than decisive normative force under certain circumstances. I offer only preliminary thoughts on the related question about the circumstances under which courts and officials may legitimately treat constitutional norms as defeasible, which is a larger and separate project. Preliminarily, I will argue that federalism norms may be treated as defeasible where the stakes for the stability of the constitutional structure are low and the non-federalism constitutional or public policy stakes are high.

A. Direct Federalism Rules

To identify direct federalism doctrines, we need to make an assumption about the basic federalism norms that they implement because the content of federalism norms is contested.43 If we assume that federalism norms specify an allocation of authority between the federal and state governments that is desirable on any of several possible criteria—e.g., fidelity to historical understandings, pragmatic value, or maximal efficiency44—then courts should formulate doctrines that preserve or restore that allocation when it is threatened. Some federalism doctrines do, in fact, appear designed to maintain or restore some sort of balance in the allocation of federal and state authority45—they are created in cases that are squarely concerned with enforcing this balance of power.46

The extent to which the Constitution actually prescribes a particular substantive allocation, much less balance in the quantum of each government’s authority, is controversial.47 But the Constitution’s clear presupposition of the durability of state governments requires some division of government power between the two levels.48 To avoid

43. See Young, Federalism Doctrine, supra note 8, at 1775–83 (highlighting the sparseness of constitutional text on federalism). Young’s taxonomy of federalism doctrine, which is more complicated, is also predicated on an assumption about the content of the underlying constitutional requirements. Id. at 1805.

44. See generally id. at 1844–48 (rehearsing federalism’s “underlying functional values”). 45. See id. at 1805 (hypothesizing this norm). 46. See id. at 1762–99 (defending a normative theory of federalism adjudication with courts

making “compensating adjustments” to correct disruptions to the “balance” of federalism). 47. See id. at 1803–1811 (canvassing the academic debate about the proper “balance” of federal

and state authority and whether such a optimal balance can be identified). 48. Excluding the Amendments, the Constitution mentions state governments in 57 provisions.

See, e.g., U.S. CONST., art. I, § 2, cls. 1–4; id. § 3 cls. 2–3, id. § 4 cl. 1, id. § 8, cls. 3, 16–17, id. § 9, cls. 1, 5–6, id. § 10, cls. 1–3; id. art. II, § 1, cls. 2–3; id. § 2, cl. 1; id. art. III, § 2, cls. 1–3; id. art. IV, § 1, § 2 cls. 1–3; id. § 3, cls. 1–2; id., § 4; id., art. VI, cls. 2–3. Even significant critics of federalism doctrine



controversies not relevant here, let us hypothesize a modest federalism norm, viz.: “The Constitution mandates the allocation of at least some authority to both the national and state governments but imposes no substantive requirements for that allocation.” This requires simply that there be both federal and state governments and suggests that federalism doctrine should prevent actions that would undermine that basic federalist structure.49 And some doctrines appear designed to identify and invalidate actions that amount to basic interference with the constitutional structure. A court could conclude that rough balance of power is a functional necessity to prevent the incremental demolition of one or the other level of government—in other words, that the abstract norm entails the more controversial one—but that is not obviously correct. There are other ways to formulate basic federalism norms,50 but these two roughly approximate the dominant judicial approaches to federalism51 will sufficiently organize our examples. The point here is to explain what courts are doing, not assess whether they have the Constitution right.

Assume that both the allocation and systemic stablity norms are valid norms in our system. Doctrines directly implementing our hypothetical federalism norms would invalidate violations by precluding (1) actions that skew a constitutionally mandatory allocation of authority between the national and state governments; (2) by actions that directly undermine the basic framework for the vertical allocation of power, for example, direct federal government interference with state governments’ internal organization; or (3) state action that interferes with either a constitutionally mandatory horizontal allocation of power among the states or the internal organization of other state governments. Indirect federalism doctrines

and practice recognize that “the Constitution makes clear that we will have some federalism in this nation.” Frank B. Cross, The Folly of Federalism, 24 CARDOZO L. REV. 1, 1 (2002).

49. Sparse constitutional specification of federalism permits only modest inferences of abstract constitutional norms. See Young, Federalism Doctrine, supra note 8, at 1775–83, 1805–06. I pitch the norm abstractly because even Young’s basic requirement of “rough balance” invites interpretive controversy. I have discussed theory-building reasons for making inferred constitutional norms as interpretively uncontroversial as possible elsewhere. See Pursley, Dormancy, supra note 20, at 533–37.

50. On the view that federalism norms require separate federal and state “spheres” of authority, see generally Edward S. Corwin, The Passing of Dual Federalism, 36 VA. L. REV. 1, 4 (1950) (arguing that judicial efforts to police separate spheres have failed); Ernest A. Young, Dual Federalism, Concurrent Jurisdiction, and the Foreign Affairs Exception, 69 GEO. WASH. L. REV. 139, 143-44 (2001). On the view that federalism norms require certain processes in federal lawmaking and little else, see generally Herbert Wechsler, The Political Safeguards of Federalism: The Role of the States in the Composition and Selection of the National Government, 54 COLUM. L. REV. 543, 543–60 (1954) (giving the canonical statement of this view); Ernest A. Young, Two Cheers for Process Federalism, 46 VILL. L. REV. 1349 (2001) (outlining process federalism views) [hereinafter Young, Two Cheers]; Bradford R. Clark, Separation of Powers as a Safeguard of Federalism, 79 TEX. L. REV. 1321 (2001).

51. See Young, Two Federalisms, supra note 1, at 15–35 (describing differing views of federalism in recent cases).



would be influenced by one of these norms but would not directly invalidate these sorts of violations.

One clear example of direct federalism doctrine can be found in the Court’s decisions expanding the scope of state sovereign immunity.52 Substantive state sovereign immunity prohibitions are based on the Eleventh Amendment and related structural “postulates that limit and control”53 the relationship between the federal and state governments. These rules “limit[] Congress’s ability to bring federal law to bear on state institutions themselves.”54 While congressional abrogation of state sovereign immunity pursuant to the Section Five power is still possible, subject to a clear statement rule I will characterize as an indirect federalism rule, below;55 abrogation pursuant to Congress’s Article I powers is now barred.56 Sovereign immunity rules appear designed to directly preclude actions that undermine what the Court has come to view as a basic structural commitment: “[States] entered the Union ‘with their sovereignty intact,’” and immunity from private lawsuits is “[a]n integral component of that ‘residuary sovereignty.’”57 Some language from these decisions suggests the Court is concerned with preventing the sort of “unanticipated intervention in the processes of [state] government” that lawsuits can cause;58 but the Court has primarily emphasized that the “purpose of state sovereign immunity is to accord States the dignity that is consistent with their status as sovereign entities.”59 That is, these rules seem aimed at “the notion of sovereignty for its own sake,”60 and courts in attempting to promote states’ dignitary interests have crafted categorical prohibitions on

52. See, e.g., Fed. Mar. Comm’n v. S.C. State Ports Auth., 535 U.S. 743, 760 (2002); Seminole Tribe v. Florida, 517 U.S. 44, 47 (1996) (Congress cannot abrogate state sovereign immunity from private damage suits when acting pursuant to its Article I powers); Alden v. Maine, 527 U.S. 706, 712 (1999) (same holding with respect to congressional authorization of private damage suits against state governments in state courts).

53. Principality of Monaco v. Mississippi, 292 U.S. 313, 322–23 (1934) (“Behind the words of the constitutional provisions are postulates which limit and control . . . .”); Seminole Tribe, 517 U.S. at 54 (suggesting that the Eleventh Amendment merely “confirms” certain background “propositions”). Sovereign immunity rules may be harder to characterize than they seem, after all, courts in these cases are at bottom interpreting the Eleventh Amendment rather than enforcing basic federalism norms. However, the purpose of the Amendment seems sufficiently connected to the purpose of federalism norms to consider them conceptually connected—a broad, abstract norm surely may have narrower, concrete instantiations in the constitutional text. See infra Part I.C.

54. Young, Two Federalisms, supra note 1, at 26–27. 55. See, e.g., Kimel v. Fla. Bd. of Regents, 528 U.S. 62, 67 (2000) (holding that abrogation of

state sovereign immunity is ineffective under the ADEA). 56. Seminole Tribe, 517 U.S. at 57–59; see Young, Two Federalisms, supra note 1, at 36 (arguing

that “the Court has [adopted] a hard rule—Congress simply may not abrogate state sovereign immunities when it acts pursuant to its Article I powers”).

57. Ports Authority, 535 U.S. at 751. 58. Alden v. Maine, 527 U.S. 706, 750 (1999). 59. Ports Authority, 535 U.S. at 760. 60. Young, Two Federalisms, supra note 1, at 27.



actions that purport to expose states to private lawsuits.61 The actual utility of sovereign immunity rules for stabilizing the system is debated.62 But the Court in recent decades has suggested that it views sovereign immunity rules as direct and important reinforcements for federalism. Doctrines directly precluding actions that expose states to suit seem like direct federalism doctrines. The now-abandoned National League of Cities doctrine, which precluded federal regulation of state government officials, institutions or employees as an intrusion on the sovereignty of “States as States,” seems like a direct federalism doctrine for the same reasons.63

Another direct federalism doctrine is the anticommandeering rule, which precludes the national government from directing state government institutions to take particular actions.64 The rule is designed to protect the basic structure of federalism; it bans a particular kind of national action, without regard to the substance of the statute or regulation at issue, because “commandeering can undermine the political safeguards that ordinarily operate to protect states” by blurring the lines of political accountability between the national and state governments.65 As the Court explained, “forcing state governments to absorb the financial burden of implementing a federal regulatory program” allows Congress to “take credit for ‘solving’ problems without having to ask their constituents to pay for the solutions with higher federal taxes” but leaves states “in the position of taking the blame for [the program’s] burdensomeness and for its defects.”66 Anti-commandeering is perhaps the most direct of the federalism doctrines; it bars one category of federal action that threatens a basic feature of the

61. See, e.g., Seminole Tribe, 517 U.S. 44; see Young, Two Federalisms, supra note 1, at 27–56; Ernest A. Young, State Sovereign Immunity and the Future of Federalism, 1999 SUP. CT. REV. 1, 27–31 (detailing the expansion of state sovereign immunity rules by the Rehnquist Court).

62. Compare, e.g., Young, Two Federalisms, supra note 1, at 154–60 (arguing that state sovereign immunity is of little practical effect)), with Alfred Hill, In Defense of Our Law of Sovereign Immunity, 42 B.C.L. REV. 485 (2000) (arguing that these cases are correctly decided); Roderick M. Hills, Jr., The Eleventh Amendment as a Curb on Bureaucratic Power, 53 STAN. L. REV. 1225 (2001) (arguing that state sovereign immunity has instrumental value).

63. Nat’l League of Cities v. Usery, 426 U.S. 833, 845 (1976), overruled by Garcia v. San Antonio Metro. Transit Auth., 469 U.S. 528, 531 (1985).

64. See New York v. United States, 505 U.S. 144, 162 (1992) (holding that Congress may not constitutionally direct state legislative action because “the preservation of the States, and the maintenance of their governments, are . . . within the design and care of the Constitution”); Printz v. United States, 521 U.S. 898, 930 (1997) (holding that the federal government may not constitutionally direct the actions of state or local government executive agencies on similar grounds).

65. Young, Two Federalisms, supra note 1, at 35 (“[T]he anticommandeering doctrine helps shore up the political safeguards of federalism by forcing the national government to internalize the costs—both fiscal and political—of its actions.”); Roderick M. Hills, Jr., The Political Economy of Cooperative Federalism: Why State Autonomy Makes Sense and “Dual Sovereignty” Doesn’t, 96 MICH. L. REV. 813, 902–03 (1998) (similar).

66. Printz, 521 U.S. at 930.



constitutional structure—the separate identities of the national and state governments.67

Some horizontal choice-of-law and interstate comity rules also may be direct federalism doctrines. These rules respond to tensions between the conventional view that states possess exclusive authority over activity inside their territories and state governments’ desires to extend their authority to extraterritorial conduct that effects their interests.68 Professor Erbsen explains that judicial involvement with horizontal conflicts between states tends to rely on three propositions that are derivable from our basic federalism norms: “states have extensive and potentially exclusive power over entities and activities physically within their territory, states may regulate based on the local effects of foreign conduct, and states may regulate domiciliaries even outside their territory.”69 Courts have crafted doctrinal rules to mediate

two tensions in this constellation of ideas. First, if states have capacity to regulate local activity with foreign effects and foreign activity with local effects, then states have overlapping capacity that undermines any claim of exclusivity and frustrates efforts to achieve local uniformity. Second, when a domiciliary of one state acts in the territory of another, both states may have inconsistent interests in regulating the activity. The unstoppable force of one state’s capacity can collide with the immovable object of the other’s capacity . . . .70

While some argue that the judicial response has been “steeped in formality but shallow in reason,” courts have developed doctrines to resolve some of these horizontal conflicts.71 Examples include rules limiting state authority to transfer title to land in another state from one domiciliary to another72 and permitting a state to refuse to enforce within its territory another state’s criminal laws even against the other state’s citizens.73 These seem like direct federalism doctrines—they are designed

67. See New York, 505 U.S. at 162 (stressing the tendency of commandeering to undermine state control of state governments); Fed. Energy Regulatory Comm’n v. Mississippi, 456 U.S. 742, 780 (1982) (O’Connor, J., dissenting) (emphasizing commandeering’s displacement of state policy choices).

68. See Erbsen, supra note 25, at 540–41 (discussing these rules as “horizontal federalism” doctrines).

69. Erbsen, supra note 25, at 562. 70. Id. at 562–63 (footnotes omitted). See generally Pennoyer v. Neff, 95 U.S. (5 Otto) 714

(1877) (state power over territory); Calder v. Jones, 465 U.S. 783 (1984) (state authority to regulate local effects of extraterritorial conduct); Curry v. McCanless, 307 U.S. 357 (1939) (state regulation of domiciliaries in other states).

71. Erbsen, supra note 25, at 563. 72. See Clarke v. Clarke, 178 U.S. 186, 193 (1900), as cited in Erbsen, supra note 25, at 563.

73. See Erbsen, supra note 25, at 563.



to reinforce a rough coequality of power between states and courts in fashioning them “rely on sweeping assertions about ‘the essential nature’ of state power and ‘constitutional barriers by which all the States are restricted within the orbits of their lawful authority.’”74 Erbsen suggests that courts’ approach to crafting interstate comity doctrine generally has been passive, perhaps conscientiously implementing a norm of federal neutrality in interstate conflicts.75 While this might mean that federalism norms are under-enforced in these contexts, existing comity rules nevertheless seem more like categorical constitutional rules than the soft, overridable requirements that characterize indirect federalism doctrine.76

The Dormant Commerce Clause doctrine, Dormant Admiralty Clause doctrine, and dormant foreign affairs doctrines also may be characterized as direct federalism rules. I have argued elsewhere that dormancy doctrines may implement a single implied constitutional preclusion against state interference with the constitutional structure.77 States may undermine stability either vertically, by interfering with national powers or interests, or horizontally, by interfering with other states. Dormancy rules safeguard the system’s stability by precluding various forms of particularly destabilizing state action—economic protectionism and interference with international relations, for example. The various dormancy doctrines include both categorical and softer, overridable prohibitions against such interference.78

These direct federalism rules tract the standard model of constitutional adjudication—both in their norm-rule-holding structure and in the courts’ treatment of federalism norms as having superordinate normative status. Indirect federalism rules depart from the standard model on both of these dimensions.

74. Id. at 564 (quoting Andrews v. Andrews, 188 U.S. 14, 34 (1903); N.Y. Life Ins. Co. v. Head, 234 U.S. 149, 161 (1914)). Erbsen notes that Congress likely could set different rules, but this does not cut against the rules’ foundation in federalism norms—the rationale is comparative institutional competence rather than constitutional mandate. See Erbsen, supra note 25, 566–67.

75. See Erbsen, supra note 25, 568–69. 76. See infra Part I.B. 77. See Pursley, Dormancy, supra note 20, at 519–23 (proposing the “State Preclusion Thesis” as

a norm grounding the various dormancy doctrines together); supra notes 21–23, 44–46 and accompanying text.

78. See Pursley, Dormancy, supra note 20, at 537–61 (discussing examples in the interstate commerce, admiralty, and foreign affairs contexts). Cf. Young, Two Federalisms, supra note 1, at 18 (discussing the “hard” vs. “soft” rule distinction and characterizing the Pike-balancing strand of the Dormant Commerce Clause doctrine as “soft”).



B. Indirect Federalism Rules

Indirect federalism doctrines like the presumption against preemption79 differ from direct federalism doctrines in that they do not directly enforce federalism norms but nevertheless are influenced by federalism. I present a non-exhaustive set of prominent examples here. Importantly, direct federalism doctrines may involve federalism norms functioning as both defeasible and decisive reasons for decision, but to avoid confusion I exclude direct federalism doctrines from the indirect category here. I want to show that this distinction between direct and indirect federalism rules cuts across conventional taxonomies of federalism doctrine and thus may advance debates in federalism theory.80

One set of indirect federalism doctrines are limitations on federal judicial power where federalism is a secondary influence on the rules. For example, abstention doctrines demonstrate that federalism considerations under some circumstances support exceptions to the norm that federal courts must exercise the jurisdiction Congress confers.81 To avoid friction between the federal and state court systems—an important vertical federalism concern82—one abstention rule requires federal courts to decline jurisdiction where duplicative state judicial or administrative proceedings are pending;83 another permits federal courts to decline jurisdiction in cases involving the constitutionality of a state law before state courts have issued a definitive construction;84 and another permits federal courts to decline jurisdiction where judicial review of state agency action “clearly involve[s]

79. See Rice, 331 U.S. at 230; supra notes 3–4 and accompanying text. 80. Professor Young’s distinction between “hard” and “soft” federalism doctrines is somewhat

similar to my direct/indirect rules distinction, but his “soft” rules category, distinguished by the overridability of the member rules, is narrower than my “indirect” rules category, which includes doctrines on which the influence of federalism norms is minor and that Young would not categorize as federalism doctrine at all. See Young, Two Federalisms, supra note 1, at 18.

81. Cohens v. Virginia, 19 U.S. 264, 404 (1821) (federal courts obligated to exercise congressionally conferred jurisdiction). On abstention doctrines, see generally Charles Allan Wright, The Abstention Doctrine Reconsidered, 37 TEX. L. REV. 815 (1959); David L. Shapiro, Jurisdiction and Discretion, 60 N.Y.U. L. REV. 543 (1985); Martin H. Redish, Abstention, Separation of Powers, and the Limits of the Judicial Function, 94 YALE L.J. 71 (1984).

82. See Wright, supra note 81, at 819 (“[I]t has been held that a federal court should refrain from exercising its jurisdiction in order to avoid needless conflict with the administration by a state of its own affairs.”); Shapiro, supra note 81, at 547 (abstention constitutes a judicial “decision not to proceed in traditional equitable principles, in principles of federalism and comity, or in principles of separation of powers”).

83. See Younger v. Harris, 401 U.S. 37, 44 (1971) (“[T]he National Government, anxious though it may be to vindicate and protect federal rights and federal interests, always endeavors to do so in ways that will not unduly interfere with the legitimate activities of the States”); Trainor v. Hernandez, 431 U.S. 434, 444 (1977) (abstention in light of parallel state administrative proceedings); Pennzoil Co. v. Texaco Co., 481 U.S. 1, 17 (1987) (extending Younger abstention to case involving two private litigants where issue involved could be raised in pending state proceeding).

84. See R.R. Comm’n of Tex. v. Pullman Co., 312 U.S. 496, 499–500 (1941).



basic problems of [state] policy.”85 Other rules counsel abstention where the case involves “a matter close to the political interests of the State,” in eminent domain proceedings.86 These rules apply regardless of substantive variations in the issues involved in particular cases; but they are soft rules in two senses: they are internal rules of judicial management that Congress could, in principle, override; and they are soft in the sense that the weight of the federalism concerns are insufficient under most circumstances to create categorical exceptions to federal jurisdiction—federal courts retain discretion on whether to abstain and conduct a case-by-case assessment of the risks of exercising jurisdiction.87 Indeed, the most open-ended abstention doctrine permits abstention under “exceptional circumstances,” including circumstances presenting judicial federalism concerns.88 Abstention rules implement—and underenforce—the basic norm that courts must exercise their jurisdiction by giving courts latitude to balance federalism consideration against instrumental concerns about the need for judicial review, allowing courts to choose which considerations are most pressing in each case. Federalism, here, is a weighty but defeasible influence on the shape of the doctrine and outcomes.

The judicial power context also provides examples of indirect federalism rules influenced by horizontal federalism concerns. A variety of doctrines are shaped in part by the need to ameliorate interstate friction that arises when multiple states pursue overlapping but conflicting regulatory interests.89 One set of such doctrines implement “individual rights tied to the multistate character of the Union and [empowering]. . .private citizens to enforce those rights in federal or state court,”90 including rights under Article IV’s Privileges and Immunities Clause against discrimination based on state citizenship;91 the right to travel under the Fourteenth Amendment

85. Burford v. Sun Oil Co., 319 U.S. 315, 332 (1943). 86. See La. Power & Light Co. v. Thibodaux, 360 U.S. 25, 28–29 (1959). 87. Thibodaux, 319 U.S. at 29 (emphasizing that “effective judicial administration” requires

“judicial discretion”); see also Colo. River Water Conservation Dist. v. United States, 424 U.S. 800, 813 (1976) (holding that courts possess discretion to determine whether a case presents “exceptional circumstances” warranting abstention).

88. Colorado River, 424 U.S. at 813. 89. See generally Erbsen, supra note 25, at 529–60. 90. Id. at 547 (citing Ex parte Young, 209 U.S. 123, 159–60 (1908) (extending federal question

jurisdiction to suits in federal court enforcing federal rights against states); Cohens v. Virginia, 19 U.S. 264, 407–10 (1821) (same, appellate jurisdiction); Howlett v. Rose, 496 U.S. 356, 367 (1990) (holding that the Supremacy Clause obligates state courts to hear federal rights claims). The discussion of these rights and their horizontal federalism implications is based in part on Erbsen, supra note 25, at 547–50.

91. See U.S. CONST. art. IV, § 2, cl. 1 (“The Citizens of each State shall be entitled to all Privileges and Immunities of Citizens in the several States Lunding v. N.Y. Tax Appeals Tribunal, 522 U.S. 287 (1998) (invalidating state statute denying tax deductions to nonresidents); Supreme Court of Va. v. Friedman, 487 U.S. 59 (1988) (invalidating state law exempting residents from the bar exam while requiring nonresidents to take it); Erbsen, supra note 25, at 548. Cf. W. & S. Life Ins. Co. v. State



Privileges and Immunities Clause;92 and the Due Process Clause right to avoid assertions of personal jurisdiction in states with which one has few or no contacts,93 among others.94 These doctrines are influenced by horizontal federalism considerations in several ways:

First, allowing individuals to protect themselves from the pitfalls of divided sovereignty imposes litigation costs on overly aggressive states that might deter abuses, avoiding interstate friction before it occurs. Second, by making unilateral overreaching more difficult, individual rights may push states toward cooperative solutions to multistate problems. Third, the availability of self-help obviates intervention by states on behalf of their citizens in disputes involving action by other states, which reduces the possibility of escalating isolated squabbles into direct interstate conflicts. Fourth, private enforcement of rights against discrimination based on place or duration of residency avoids creating a class of “stateless” citizens who upon leaving one state lack the protection of another, which might make them a source of political tension between states vying to exclude them. Finally, requiring states to treat each others’ citizens approximately as well as they treat their own helps establish a national identity that might override or mitigate regional parochialism.95

The transformation and expansion of due process restrictions on state court assertions of personal jurisdiction over the last century in response to changes in interstate travel and commerce suggest that the weight of federalism’s influence on these doctrines relative to other considerations varies with the circumstances.96 Federalism concerns arguably were outweighed by instrumental considerations about the prospective workability of doctrine in the assimilation of quasi-in rem jurisdiction to

Bd. of Equalization, 451 U.S. 648, 657 (1981) (recognizing similar right under Equal Protection Clause subject to rational basis review).

92. See Saenz v. Roe, 526 U.S. 489, 502–03 (1999). 93. See, e.g., Ins. Corp of Ir. v. Compagnie des Bauxites de Guinee, 456 U.S. 694, 702 (1982);

Int’l Shoe Co. v. Washington, 326 U.S. 310 (1945) (establishing due process rule requiring “minimum contacts” between defendant and forum state before personal jurisdiction may permissibly be asserted).

94. See Erbsen, supra note 25, at 548 (discussing “the right under an ill-defined constellation of clauses to be free from the extraterritorial operation of state laws,” “rights with a horizontal dimension” created by the “Double Jeopardy and Contracts Clauses,” and privately enforceable rights under the dormant Commerce Clause).

95. Id. at 549–50 (footnotes omitted). 96. See generally Pennoyer v. Neff, 95 U.S. (5 Otto.) 714 (1877); Int’l Shoe, 326 U.S. 310;

Shaffer v. Heitner, 433 U.S. 186 (1977); Hanson v. Denckla, 357 U.S. 235 (1958); World-Wide Volkswagen Corp. v. Woodson, 444 U.S. 286 (1980); Bauxites, 456 U.S. 694; Burger King Corp. v. Rudzewicz, 471 U.S. 462 (1985); Asahi Metal Indus. Co. v. Superior Court, 480 U.S. 102 (1987) (plurality op.).



the more restrictive Due Process standard applied to assertions of in personam jurisdiction97—the benefits of streamlined doctrine trumped state governments’ interest in having their own courts hear certain cases involving in-state property.

The doctrine governing the power of federal courts to craft federal common law rules of decision is influenced, at least in part, by similar horizontal federalism considerations.98 Federal common lawmaking is controversial—the most obvious objection is that federal courts are not chartered to legislate, a separation-of-powers critique.99 But perhaps the most palatable justification for crafting rules of federal common law is that the practice allows federal courts to supply neutral rules for resolving interstate disputes or disputes whose resolution no individual state has the capacity to affect alone.100 If the substantive law of the states involved favors parochial interests, applying that law might ignite interstate tension and, perhaps, make it difficult to achieve a satisfactory resolution.101 Thus, courts have held that federal common law legitimately may be formulated and applied in “border disputes[,]. . . actions involving interstate pollution or the downstream effects of upstream water uses[,] . . . and disputes about intangible property within multistate contacts.”102 This collective-action-problem-solving rationale for federal common lawmaking is widely accepted and viewed as defensible even by some who are critics of federal common law in other areas.103 Courts balance the federalism considerations against the separation of powers concerns: The former may outweigh the latter and justify judge-made rules in interstate disputes; but a prominent

97. See Shaffer v. Heitner, 433 U.S. 186 (1977) (holding that assertions of personal jurisdiction must satisfy International Shoe’s “minimum contacts” standard regardless of defendant’s property in the state).

98. See Erbsen, supra note 25, at 558–60 (discussing federal common law from the horizontal federalism perspective); see also id. at 556–67, 556 n.232 (noting that the question whether federal common law is permissible is technically about the meaning of the “judicial power” language of Article III).

99. For an overview of the debate, see generally Monaghan, supra note 39, at 758–65; Ernest A. Young, Preemption and Federal Common Law, 83 NOTRE DAME L. REV. 1639, 1661 (2008) [hereinafter Young, Common Law].

100. See Jay Tidmarsh & Brian J. Murray, A Theory of Federal Common Law, 100 NW. U. L. REV. 585, 631 (2006) (arguing that federal common lawmaking is a corrective to state self-dealing).

101. See Erbsen, supra note 25, at 555 (arguing that interstate tensions can arise, and federal common law might be the proper palliative, “when states attempt to exercise dominion over other states, to externalize costs onto other states, or to overreach their borders by regulating activity in other states”).

102. Id. at 45; see, e.g., Illinois v. City of Milwaukee, 406 U.S. 91 (1972) (downstream pollution disputes); Texas v. New Jersey, 379 U.S. 674 (1965) (intangible property subject to multistate interests); New Jersey v. New York, 283 U.S. 336 (1931) (externalities); Cissna v. Tennessee, 246 U.S. 289 (1918) (border disputes).

103. See, e.g., Bradford R. Clark, Federal Common Law: A Structural Reinterpretation, 144 U. PA. L. REV. 1245, 1322 (1996) (articulating the state incompetence rationale for federal common law-making); Tidmarsh & Murray, supra note 100, at 631–35 (noting this view).



judicial view after Erie is that federal common law otherwise should be restricted to interstitial applications, suggesting that separation-of-powers considerations often are viewed as weightier.104 Federal common law rules tend to displace the substantive state law that would have applied under conventional choice-of-law principles; this introduces an overridable vertical federalism concern that may contribute additional support for decisions to forego federal common law-making. Federalism considerations thus appear to be defeasible reasons bearing on the rule governing the permissibility of federal common law-making, even though that standard is directed at the scope of federal judicial power.

Some constitutional rights doctrine also displays indirect federalism characteristics.105 Existing doctrine makes every constitutional right overridable by sufficiently strong (and fairly pursued) state interests.106 Consider the Equal Protection doctrine’s famous tiers of scrutiny, which provide a formulaic tool for courts to assess the state interests at stake in alleged rights violations.107 The routine manner in which these rules are applied today obscures federalism’s continuing influence—judicial decisions about the standard applicable to state action (strict scrutiny108 to rational basis review109) are influenced by both the nature of the rights

104. Erie R.R. Co. v. Tompkins, 304 U.S. 64, 78 (1938); see also City of Milwaukee v. Illinois, 451 U.S. 304, 312 (1981) (“Federal courts, unlike state courts, are not general common-law courts and do not possess a general power to develop and apply their own rules of decision.”); United States v. Kimbell Foods, Inc., 440 U.S. 715, 727 (1979) (emphasizing the interstitiality of federal common law); Curtis A. Bradley, Jack L. Goldsmith & David H. Moore, Sosa, Customary International Law, and the Continuing Relevance of Erie, 120 HARV. L. REV. 869, 880 (2007) (same). Cf. Thomas W. Merrill, The Common Law Powers of Federal Courts, 52 U. CHI. L. REV. 1, 47 (1985). But see Martha A. Field, Sources of Law: The Scope of Federal Common Law, 99 HARV. L. REV. 881, 887 (1986) (arguing that federal common lawmaking must be authorized by statute).

105. See Young, Federalism Doctrine, supra note 8, at 1793–94 (“The potential of individual rights decisions to restrict the autonomy of state governments is well understood . . . . It is now commonplace to think of a case like Lawrence v. Texas . . . as raising significant issues of federalism.” (footnotes omitted) (also citing Duncan v. Louisiana, 391 U.S. 145, 172–73 (1968) (Harlan, J., dissenting) (arguing that incorporation of federal constitutional rights has “fasten[ed] on the States federal notions of criminal justice”))); Lynn A. Baker & Ernest A. Young, Federalism and the Double Standard of Judicial Review, 51 DUKE L.J. 75, 157–59 (2001) (similar claim regarding the federalism implications of free speech and association rights doctrines).

106. See, e.g., Roe v. Wade, 410 U.S. 113, 154 (1973) (explaining that “the right of personal privacy . . . is not unqualified and must be considered against important state interests”).

107. See generally U.S. CONST., amend. XIV, §1 (Equal Protection Clause); Reva B. Siegel, Equality Talk: Antisubordination and Anticlassification Values in Constitutional Struggles over Brown, 117 HARV. L. REV. 1470, 1495–1499 (2004) (canvassing interpretations of the Equal Protection clause).

108. See, e.g., Adarand Constructors, Inc. v. Pena, 515 U.S. 200, 227 (1995) (explaining that strict scrutiny requires that classifications be invalidated unless “narrowly tailored . . . to achieve compelling governmental interests”); In re Griffiths, 413 U.S. 717, 721–22 (1973) (giving a canonical formulation).

109. See, e.g., FCC v. Beach Commc’ns, Inc., 508 U.S. 307, 313 (1993) (explaining that certain classifications are permissible unless no “conceivable state of facts … could provide a rational basis for the classification.”); Bd. of Trs. of the Univ. of Ala. v. Garrett, 531 U.S. 356, 366–67 (2001) (applying rational basis review).



violation alleged and its importance for state government functioning (and thus to system stability).110 For example, in San Antonio Independent School District v. Rodriguez,111 the Court held that an Equal Protection challenge to a state public school funding regime based on local property taxes should be evaluated under the lenient rational basis standard because tax policy requires particularized “expertise” and “familiarity with local problems,” suggesting that judicial deference to state decisions will minimize adjudicatory errors.112 Selecting the deferential approach addressed federalism concerns about both the fact that “systems of financing public education presently in existence in virtually every state” were at stake,113 and about how a more “inflexible constitutional restraint[]” like strict scrutiny might inhibit state policy experimentation.114 In short, conventional federalism considerations—preserving state policymaking prerogatives, allowing state-by-state regulatory variation, and limiting federal government intervention in state affairs—influence constitutional rights doctrine because the Court appears to view the practice of accounting for federalism as important for generating optimal adjudicatory outcomes. The same federalism concerns might be discounted if the facts change such that different doctrine is more likely to result in right decisions.

Takings jurisprudence presents another example.115 In holding that states may take private property for private development, and thus adding a controversial feature to an already complex doctrine, the Court expressly considered federalism issues, including the extent to which the constitutional question is one that states are well situated to decide and to which variations in local conditions are relevant:

110. See Sager, supra note 10, at 1218–19. For another example of state interests shaping equal protection doctrine, note that strict scrutiny is justified in part by the historical observation that certain kinds of classifications frequently have been used for illegitimate reasons, and seldom for legitimate ones”—e.g., racial classifications—so that a stricter decision rule likely “will invalidate many unconstitutional laws and very few legitimate ones.” Roosevelt, supra note 28, at 1663–64; Johnson v. California, 543 U.S. 499, 505 (2005) (“Racial classifications raise special fears that they are motivated by an invidious purpose.”).

111. 411 U.S. 1 (1973). 112. Rodriguez, 411 U.S. at 41. 113. See id. at 44. 114. See id. at 43; Sager, supra note 10, at 1218–19 (discussing Rodriguez and the federalism

considerations that influenced the outcome). 115. See generally Timothy Sandefur, The “Backlash” So Far: Will Americans Get Meaningful

Eminent Domain Reform?, 2006 MICH. ST. L. REV. 709; William A. Fletcher, Kelo, Lingle, and San Remo Hotel: Takings Law Now Belongs to the States, 46 SANTA CLARA L. REV. 767 (2006); Owen Lipsett, Comment, The Failure of Federalism: Does Competitive Federalism Actually Protect Individual Rights?, 10 U. PA. J. CONST. L. 643, 655–59 (2008) (all examining the federalism component of takings doctrine).



[O]ur jurisprudence has recognized that the needs of society have varied between different parts of the Nation, just as they have evolved over time in response to changed circumstances. Our earliest cases in particular embodied a strong theme of federalism, emphasizing the ‘great respect’ that we owe to state legislatures and state courts in discerning local public needs. For more than a century, our public use jurisprudence has wisely eschewed rigid formulas and intrusive scrutiny in favor of affording legislatures broad latitude in determining what public needs justify the use of the takings power.116

The takings context also displays a parallel to the judicial power context—the Court has adopted an abstention rule precluding lower federal courts from hearing takings cases initiated in state court,117 in part to account for state interests in having state courts construe the challenged actions in the light of local considerations.118 Both rules serve federalism concerns by preserving state adjudicatory power on institutional competence grounds; but the implication is that instrumental developments could outweigh federalism concerns and drive courts to a different rule if they alter perceptions of state capacity, suggest that local conditions are unimportant, or show that significant federal interests are also at stake. Here, federalism, again, looks like a defeasible reason in doctrine-making.

Federalism also influences the formulation of modern doctrine concerning the recognition of unenumerated constitutional rights. Courts view federalism itself as directed to the protection of individual rights.119 Judicial deliberation about recognizing constitutional rights to abortion,120 physician assisted suicide,121 the use of medical marijuana,122 same-sex

116. Kelo v. City of New London, 545 U.S. 469, 482–83 (2005); see also Lingle v. Chevron U.S.A., Inc., 544 U.S. 528, 548 (2005) (affirming California property rental restrictions and holding that Takings Clause restricts only physical takings, total regulatory takings, and land-use exactions onerous enough to amount to physical takings).

117. See San Remo Hotel, L.P. v. City & Cnty. of San Francisco, 545 U.S. 323, 347 (2005); see also Williamson Cnty. Reg’l Planning Comm’n v. Hamilton Bank, 473 U.S. 172, 186 (1985) (holding that takings claims against state governments must ripen in state court).

118. See Fletcher, supra note 115, at 776 (arguing that recent takings doctrine “relegat[es] takings issues to the political and legal judgments of the states”).

119. See, e.g., New York v. United States, 505 U.S. 144, 181 (1992) (“[T]he Constitution divides authority between federal and state governments for the protection of individuals. . . . [F]ederalism secures to citizens the liberties that derive from the diffusion of sovereign power.”).

120. See, e.g., Roe v. Wade, 410 U.S. 113, 154 (1973) (recognizing that the constitutional right to privacy includes the right to abortion).

121. See, e.g., Washington v. Glucksberg, 521 U.S. 702, 735 (1997) (upholding Washington’s prohibition on physician-assisted suicide over constitutional challenge).

122. Cf. Gonzales v. Raich, 545 U.S. 1 (2005) (holding that Controlled Substance Act regulation precluding medical marijuana use even as permitted by state law was constitutional despite federalism



marriage,123 and so forth often turns in part on the extent to which states collectively recognize or repudiate the relevant rights.124 The Court has also considered variations in states’ interests in regulating the challenged conduct—in the physician assisted suicide cases, for example, the Court has relied on the variability of state interests, among other factors, to determine whether doctrine implementing the Due Process Clause should categorically permit or bar the practice:125

Whether or not Missouri’s clear and convincing evidence requirement [for patient consent to terminate life support] comports with the United States Constitution depends in part on what interests the State may properly seek to protect in this situation. Missouri relies on its interest in the protection and preservation of human life . . . . As a general matter, the States—indeed, all civilized nations—demonstrate their commitment to life by treating homicide as a serious crime. Moreover, the majority of States in this country have laws imposing criminal penalties on one who assists another to commit suicide. . . . . . . We believe Missouri may legitimately seek to safeguard the personal element of this choice through the imposition of heightened evidentiary requirements. It cannot be disputed that the

concerns); Ernest A. Young, Just Blowing Smoke? Politics, Doctrine, and the Federalist Revival After Gonzales v. Raich, 2005 SUP. CT. REV. 1 (assessing the federalism implications of Raich).

123. See Lawrence v. Texas, 539 U.S. 558, 578 (2003) (stressing that striking down anti-sodomy laws “does not involve whether the government must give formal recognition to any relationship that homosexual persons seek to enter”). State courts have begun developing rights for same-sex couples based on a variety of legal provisions. See, e.g., Strauss v. Horton, 207 P.3d 48 (Cal. 2009) (recognizing same-sex partners entitled to equal protection by California Constitution but rejecting right to formal “marriage”); Martinez v. County of Monroe, 850 N.Y.S.2d 740 (N.Y. App. Div. 2008) (holding that New York law requires recognizing out-of-state same-sex marriages).

124. In identifying rights fundamental enough to be considered substantive guarantees of the Due Process Clauses, the Court typically looks for rights “deeply rooted in this Nation’s history, [practices], and tradition,” which are “crucial ‘guideposts for responsible decisionmaking.’” Glucksberg, 521 U.S. at 721. The Court regards state legislatures “the primary and most reliable indication of . . . consensus.” Id. at 711 (citing Stanford v. Kentucky, 492 U.S. 361, 373 (1989)). Thus, the Court invalidated state anti-sodomy laws based, in part, on an “emerging awareness that liberty gives substantial protection to adult persons in deciding how to conduct their private lives in matters pertaining to sex.” Lawrence, 539 U.S. at 571–72. See generally Corinna Barrett Lain, The Unexceptionalism of “Evolving Standards,” 57 UCLA L. REV. 365, 419 (2009) (noting that “[a]cross a stunning variety of civil liberties contexts, the Court routinely—and explicitly—decides constitutional protection based on whether a majority of states agree with it”).

125. Cruzan v. Dir., Mo. Dep’t of Health, 497 U.S. 261, 285–86 (1990). Compare, e.g., Cruzan, 497 U.S. at 281–82; (upholding Missouri state evidentiary requirement for patient consent), and Glucksberg, 521 U.S. at 717–18 (upholding Washington ban on physician assisted suicide), with Vacco v. Quill, 521 U.S. 793, 797 (1997) (upholding New York law permitting patients to refuse life-saving medical measures but prohibiting physician-assisted suicide), and Gonzales v. Oregon, 546 U.S. 243, 267 (2006) (striking down Attorney General ruling inconsistent with Oregon law permitting physician-assisted suicide).



Due Process Clause protects an interest in life as well as an interest in refusing life-sustaining medical treatment . . . A State is entitled to guard against potential abuses in such situations.126

On most interpretive theories, the existence of a constitutional right has more to do with the proper meaning of the relevant provision than with the right’s potential to interfere with state regulatory choices. Federalism here, too, is at best an indirect influence on doctrine.

The most familiar indirect federalism doctrines are canons of statutory interpretation. These are, most directly, about implementing legislative supremacy by accurately ascertaining the meaning of federal enactments.127 As with constitutional interpretation, there is debate about how to determine “accuracy” or “correctness” in statutory interpretation—the clash is primarily between purposivists who argue that interpretation should be about identifying what the legislature intended and textualists who urge fidelity, above all, to what the legislature actually said.128 Interpretive canons influenced by federalism norms include the presumption against preemption, a state-autonomy-influenced rule which, at least for federal statutes that purport to reach areas of “traditional state interest,”129 requires courts to avoid construing federal statutes to preempt state law absent a clear indication of congressional intent to preempt.130 The rule of Gregory v. Ashcroft requires courts to avoid construing federal statutes to impose obligations or liabilities on state government institutions or officials in the absence of clear statutory language131—a presumption aimed at the same sort of intrusions into state sovereignty (or autonomy, depending on one’s perspective132) that are the direct objects of the National League of Cities and anticommandeering doctrines.133 In Pennhurst State School and Hospital v. Halderman,134 the Court established a rule against interpreting federal statutes to impose conditions on federal funding to states without an

126. Cruzan, 479 U.S. at 280–81. 127. See Young, Two Federalisms, supra note 1, at 47 (“All of these rules are, of course, rules of

statutory construction.”); Gardbaum, supra note 3, at 768 (noting that “preemption has largely been ignored by constitutional law scholars” who regard it as an issue of mere statutory construction).

128. See, e.g., Manning, supra note 7, at 2013–20 (setting out the basics of strict textualism). 129. See, e.g., United States v. Locke, 529 U.S. 89, 108 (2000); Crosby v. Nat’l Foreign Trade

Council, 530 U.S. 363, 374 n.8 (2000) (both failing, without explanation, to apply the presumption against preemption in standard preemption cases). Cf. Davis, supra note 5, at 1219 (noting that these and other omissions lack clear grounds).

130. Rice v. Santa Fe Elevator Corp., 331 U.S. 218, 230 (1947). 131. Gregory v. Ashcroft, 501 U.S. 452 (1991); Eskridge & Frickey, supra note 4, at 623–24. 132. See generally Young, Two Federalisms, supra note 1, at 63–65 (stressing the “[o]verriding

[s]ignificance of [state government] [a]utonomy” compared to the relatively lesser significance of state “sovereignty” in making federalism doctrine).

133. See supra notes 63–67 and accompanying text. 134. 451 U.S. 1 (1981).



unambiguous statement from Congress.135 The Penhurst rule is influenced by federalism concerns, specifically concerns about voluntary state government decisionmaking: “the crucial inquiry,” according to the Court, is “whether Congress spoke so clearly that we can fairly say that the State could make an informed choice” to accept the spending condition.136 Other federalism-influenced statutory interpretation rules require clear statements from Congress before federal statues will be read to abrogate state sovereign immunity,137 to apply criminal liability to state or local government officials,138 or to press the outer limits of Congress commerce power to the potential derogation of reserved state regulatory authority.139

These federalism-backed interpretive canons are somewhat detached from conventional accounts of the purpose of interpretive rules.140 They are in some sense “normative canons”—pushing statutory interpretation toward what the court thinks is a normatively desirable outcome even if the result is not the most accurate assessment of the statute’s meaning.141 Normative canons are for this reason controversial, but for present purposes the debate about their legitimacy is of secondary importance.142 For now, it is enough to observe that courts employ several canons that are influenced by federalism considerations. In this sense, federalism’s influence on these doctrines is stronger than on perhaps any other indirect doctrines. Nevertheless, the federalism-influenced interpretive canons are paradigmatic “soft” federalism doctrines:143 They do not invalidate actions that violate federalism norms; instead, they interpret away federalism-

135. Id. at 17; see Eskridge & Frickey, supra note 4, at 619–20 (discussing Pennhurst). 136. Pennhurst, 451 U.S. at 24–25. 137. See, e.g., Atascadero State Hosp. v. Scanlon, 473 U.S. 234, 240 (1985); see Eskridge &

Frickey, supra note 4, at 621–22 (discussing these canons). 138. McNally v. United States, 483 U.S. 350 (1987); McCormick v. United States, 500 U.S. 257

(1991); Eskridge & Frickey; supra note 4, at 626–27. 139. Jones v. United States, 529 U.S. 848, 859 (2000); Solid Waste Agency v. U.S. Army Corps.

of Eng’rs, 531 U.S. 159, 174 (2001); see Young, Two Federalisms, supra note 1, at 37–38 (discussing these rules).

140. See Young, Two Federalisms, supra note 1, at 47 (“Clear statement rules matter only when they cause a court to pick an interpretation of a statute other than the one they would have picked absent the rule.”). Cf. Frederick Schauer, Ashwander Revisited, 1999 SUP. CT. REV. 71, 87 (noting a similar dynamic with respect t to constitutional avoidance canons).

141. See Young, Two Cheers, supra note 50, at 1387 (distinguishing “‘descriptive’ canons, which embody predictive judgments about how Congress most likely would have wanted certain kinds of statutory ambiguity resolved; [from] ‘normative’ canons, which seek to protect certain kinds of public values whether or not Congress probably intended to do so”). See generally Young, Resistance Norms, supra note 11; Sunstein, supra note 11 (both canvassing interpretive canons).

142. See Young, Two Cheers, supra note 50, at 1388–89 (canvassing objections); Richard A. Posner, Statutory Interpretation--in the Classroom and in the Courtroom, 50 U. CHI. L. REV. 800, 816 (1983) (arguing that normative canons function like constitutional constraints and are of questionable legitimacy).

143. See Young, Two Federalisms, supra note 1, at 15–18 (characterizing substantive canons as federalism doctrine).



related concerns. Some cue Congress to enact clearer language that could push courts to decide whether direct federalism violations have occurred—canons of constitutional avoidance function this way144—but most of these rules are truly soft—their shaping federalism concerns fully defeasible—in the sense that enacting clearer language is all Congress has to do to resolve the Court’s federalism concerns even if the statute, as a result, intrudes deeply into state prerogatives. The presumption against preemption is an example of this latter kind of rule.145

Interpretive canons do substantial federalism-related work; and that’s not difficult to understand.146 The Supreme Court endorsed the “political safeguards of federalism” in Garcia, announcing that “the principal means chosen by the Framers to ensure the role of the States in the federal system lies in the structure of the Federal Government itself.”147 Process-reinforcing interpretive rules are the kind of federalism doctrine that advocates of the political safeguards view tend to endorse:148 “Rules requiring that Congress speak clearly when it intrudes on state prerogatives enhance the political and institutional checks on such intrusions . . . .”149 Even if our hypothetical federalism norms are wrong and the Constitution in fact requires only the proper functioning of the national lawmaking process, these rules remain indirect rules because they do not invalidate laws that result from deficient processes. The force of the federalism concerns that influence most of these rules can be overcome by ordinary lawmaking whether or not the results of that lawmaking process actually violate underlying federalism norms. Thus, we must confront the conceptual deeper question: How can federalism norms influence judicial doctrine in this defeasible way, different from the decisive force of constitutional norms on the standard model?

144. See Young, Resistance Norms, supra note 11; Pursley, Preemption Decisions, supra note 9. 145. See supra notes 3–5 and accompanying text. One might quibble with my categorization here,

as several of these interpretive canons are so strongly influenced by considerations of federalism that some might characterize them as directly “about” federalism. They may, in this sense, be another category of hard cases. See infra Part I.C. If we were defining the categories according to various doctrines’ importance for safeguarding some “‘balance” of federalism, these interpretive canons might well be the most important of all federalism doctrines. See Young, Two Federalisms, supra note 1, at 130–34. But my goal to analyze federalism norms’ role effects on constitutional doctrine—like the norm instructing courts to choose the most accurate interpretation of a statute, however accuracy is measured—requires categorizing these canons as indirect rules.

146. See Young, Two Federalisms, supra note 1, at 130–34 (explaining the centrality of federalism canons).

147. Garcia v. San Antonio Metro. Transit Auth., 469 U.S. 528, 550 (1985). 148. See Young, Two Federalisms, supra note 1, at 19–20 (noting the “obvious affinity between

process federalism and soft limits on federal power, particularly in the case of clear statement canons of statutory construction”).

149. Id. at 19.



These rules vary in the extent to they are influenced by federalism. While the presumption against preemption imposes only a requirement that Congress make clear its preemptive intent, the Gregory rule and the canon regarding abrogation of state sovereign immunity are “super strong,” requiring clear language before the adverse effect on state governments will be read into the statute.150 The Penhurst “unambiguous” statement requirement is somewhere in between.151 The legislative cost of demonstrating intent may be lower than the cost of shepherding clear abrogation language through the legislative process; these rules thus impose federalism-influenced constraints of varying force on congressional practice.152 Their strengths vary with the perceived importance of the underlying federalism concern that is responsible for the rule; the super strong rules relate to federalism concerns like interference with internal state government ordering and sovereign immunity that are similar in kind to those that have prompted direct federalism doctrines.153 The weaker canons are influenced by federalism concerns—like the effects on state regulatory autonomy of federal preemption of state law—that have been of less concern to the courts.154

C. Hard Cases

Several doctrines resist easy categorization on the direct/indirect federalism distinction. Some are hard to categorize because the formative decisions appear grounded on non-federalism norms but contain extensive discussions of federalism such that federalism’s apparent influence makes it hard to resist thinking that they are “about” federalism. The best example may be Erie,155 where the Court rejected the longstanding practice of applying federal common law rules in diversity cases and held that substantive state-law rules must apply.156 The Justices in Erie articulated several rationales for the decision, including a statutory construction holding that the federal Rules of Decision Act requires the application of state common law in diversity cases absent an applicable federal statutory

150. Eskridge & Frickey, supra note 4, at 621–22 (abrogation canon); id. at 623–24 (Gregory rule).

151. Id. at 619–20 (discussing the Pennhurst canon). 152. See Young, Two Federalisms, supra note 1, at 19–20 (discussing the costs that clear

statement requirements add to the legislative process); James J. Brudney, Congressional Commentary on Judicial Interpretation of Statutes: Idle Chatter or Telling Response?, 93 MICH. L. REV. 1, 30 (1994) (similar).

153. See supra Part I.A. 154. See Young, Two Federalisms, supra note 1, at 134–36.

155. Erie R. Co. v. Tompkins, 304 U.S. 64 (1938), overruling Swift v. Tyson, 41 U.S. 1 (1842). 156. See Erie, 304 U.S. at 78 (declaring that “[t]here is no general federal common law”).



or regulatory rule of decision;157 and a holding apparently based one of several constitutional rationales mentioned, including federalism and equal protection concerns. The Court did not specify the precise constitutional violation.158 Federalism considerations weighed in the decision, but debate continues about the ultimate grounding of the decision.159 Erie at least states an indirect federalism rule—whatever the ground for decision was, it was clearly influenced by federalism considerations. The Court’s lengthy discussion of federalism160 and the growth of an entire line of jurisprudence and scholarship concerned with “judicial federalism” taking Erie as its touchstone create the impression that Erie was an important federalism case.161

Another possible constitutional basis for Erie is that “Congress has no power to declare substantive rules of common law applicable in a state whether they be local in their nature or ‘general’” and “no clause in the Constitution purports to confer such a power on the federal courts.”162 If this is the ground of decision, then Erie implements a constitutional limitation on federal power—specifically, federal judicial power—that is influenced by federalism considerations and thus similar to other judicial

157. See id. at 91 (Reed, J., concurring in part); Craig Green, Repressing Erie’s Myth, 96 CAL. L. REV. 595, 602 (2008) (noting that this basis alone could have justified overruling Swift). Courts and casebooks sometimes characterize Erie as resting on a reinterpretation of the RDA. See, e.g., Gasperini v. Ctr. for Humanities, 518 U.S. 427 (1996) (characterizing Erie’s holding as predicated on reinterpreting the Rules of Decision Act); JACK H. FRIEDENTHAL, ARTHUR R. MILLER, ET AL., CIVIL

PROCEDURE: CASES AND MATERIALS 415 (10th ed. 2010) (“The Constitutional discussion in Erie is sometimes referred to as ‘dicta.’ Is Justice Brandeis’s reference to the Constitution merely a way of bolstering his interpretation of the Rules of Decision Act?”); RICHARD FALLON, DANIEL MELTZER &

ROBERT SHAPIRO, HART & WECHSLER’S THE FEDERAL COURTS AND THE FEDERAL SYSTEM 563–64 (5th ed. 2009) [hereinafter HART & WECHSLER] 563–64 (noting debate over the constitutional basis for Erie).

158. See Erie, 304 U.S. at 74–75, 78 (“Swift . . . made rights enjoyed under the unwritten ‘general law’ vary according to whether enforcement was sought in the state or federal court,” rendering “impossible the equal protection of the law.”).

159. This lack of a clear explanation of the specific constitutional rationale for decision has created a longstanding debate over the nature of Erie’s actual holding. See generally EDWARD A. PURCELL, JR., BRANDEIS AND THE PROGRESSIVE CONSTITUTION: ERIE, THE JUDICIAL POWER, AND THE

POLITICS OF THE FEDERAL COURTS IN TWENTIETH CENTURY AMERICA 178–79 (2000) (expressing skepticism about the likelihood that the decision actually rested on Brandeis’s “oblique” and “reluctant” states’ rights rhetoric); HART & WECHSLER, supra note 157, at 563–64 (noting this debate); see also Green, supra note 157, at 610–11 (arguing that “Erie’s states rights argument is of questionable merit” and that “modern states-rights federalism cannot support Erie’s constitutional holding and premodern arguments about dual sovereignty can only undermine Erie’s present status”); John Hart Ely, The Irrepressible Myth of Erie, 87 HARV. L. REV. 697, 702 (1974) (rejecting the federalism rationale for Erie).

160. See generally Erie, 304 U.S. at 71–80. 161. See, e.g., Hanna v. Plummer, 380 U.S. 460, 474 (1965) (Harlan, J., concurring) (calling Erie

“a cornerstone of our federalism”). 162. Erie, 304 U.S. at 78.



power rules. That would, again, suggest that Erie is properly an indirect federalism case.163

Commerce Clause doctrine is also hard to categorize. Commentators often characterize recent decisions narrowing the commerce power as part of the Rehnquist Court’s “Federalist Revival.”164 In United States v. Lopez and United States v. Morrison, the Court held that the Commerce Clause does not empower Congress to regulate “non-commercial” activity of primarily intrastate effect, citing among other things, the potential for a broader rule to constrict state authority.165 That may seem like a straightforward interpretation of the text of the Commerce Clause; and on that view Lopez and Morrison are federalism-influenced legislative power decisions, and, thus, establish at most indirect federalism doctrine. The Court discussed the federalism benefits of limiting federal power at length.166 But it did not specify whether the language of the Commerce Clause, or, instead, federalism norms were the constitutional basis for the limitation. The Lopez Court discussed examples of internal “limitations on the commerce power that are inherent in the very language of the Commerce Clause”167 and external limitations168 that exist because “certain categories of activity such as ‘production,’ ‘manufacturing,’ and ‘mining’ were within the province of state governments and thus were beyond” the commerce power.169 Nobody doubts that the results in Lopez and Morrison were shaped by federalism concerns; but the tight relationship between the

163. Recent scholarship posits a separation-of-powers basis for Erie, though it’s not firmly tethered to opinion. See generally Clark, supra note 50.

164. See, e.g., Young, Two Federalisms, supra note 1, at 15–18 (discussing Lopez and Morrison); see also Vicki C. Jackson, Federalism and the Uses and Limits of Law, Printz and Principle?, 111 HARV. L. REV. 2180, 2213 (1998) (branding this “federalist revival”).

165. United States v. Lopez, 514 U.S. 549, 567 (1995); United States v. Morrison, 529 U.S. 598, 617 (2000). For some, these are quintessential federalism cases. See, e.g., Viet D. Dinh, Reassessing the Law of Preemption, 88 GEO. L.J. 2085, 2117–18 (2000). But see Young, Two Federalisms, supra note 1, at 135 (arguing that “Lopez and Morrison are likely to have only limited practical significance”). Even Professor Young does not deny the importance of limiting national power for federalism—indeed, he notes, “develop[ing] meaningful limits under the Commerce Clause” is a problem that has “dominated the history of federalism jurisprudence.” Young, Two Federalisms, supra note 1, at 136.

166. Lopez, 514 U.S. at 564, 567; see Morrison, 529 U.S. at 617–19 (noting that the “Constitution requires a distinction between what is truly national and what is truly local” and that “[t]he regulation and punishment of intrastate violence that is not directed at the instrumentalities, channels or goods involved in interstate commerce has always been the province of the States”).

167. Lopez, 514 U.S. at 552–58. Internal limitations are those that arise from the terms of the grant of power itself; external limitations are those that arise from other constitutional provisions, like the Due Process Clause. See generally LAURENCE H. TRIBE, AMERICAN CONSTITUTIONAL LAW 297 (2d ed. 1988) (distinguishing internal and external limitations); see, e.g., Perez v. United States, 402 U.S. 146 (1971) (internal limitation); Garcia v. San Antonio Metro. Transit Auth., 469 U.S. 528 (1985) (external Tenth Amendment-based limitation); Lane County v. Oregon, 74 U.S. 71 (1868) (similar).

168. The result of applying even the internal limitations, after all, is to preserve portions of “the province of state governments,” since that province is simply defined as whatever is not subject to federal regulation.

169. Lopez, 514 U.S. at 554.



scope of federal power and that of state power makes it difficult to distinguish internal limitations from external federalism-based external limitations in enumerated powers cases. Thus, it is hard to say whether these cases articulate a direct or indirect federalism rule.170

Some commentators regard this sort of limitation on federal power as arising from federalism norms.171 After all, the Tenth Amendment172 defines state power in relation to federal power—the greater the scope of federal power, the potentially smaller the scope of state power.173 A claim that Lopez and Morrison are direct federalism cases might rely on one of two views: One might think, first, that limiting federal power is intrinsically important for federalism because it creates space for states to act without the possibility of federal interference even if the limitations do not arise directly from constitutional federalism norms.174 Second, one might argue that constitutional federalism norms themselves directly limit national government power in certain areas. One way to defend this second argument is to take a strong view of the relationship between federal and state power established by the Tenth Amendment—if every kind of authority not delegated to the national government is reserved to the states, then all rules defining the scope of federal government power are federalism rules by necessary implication. As they define federal power, they also define and protect the scope of state power.175 If every

170. See New York v. United States, 505 U.S. 144, 155 (1992) (“These questions can be viewed in either of two ways. In some cases the Court inquired whether an Act of Congress is authorized by one of the powers delegated to Congress in Article I of the Constitution. In other cases the Court has sought to determine whether an Act of Congress invades the province of state sovereignty reserved by the Tenth Amendment. In a case like these, involving the division of authority between federal and state governments, the two inquiries are mirror images of each other.”).

171. See, e.g., Young, Two Federalisms, supra note 1, at 33–35 (characterizing recent enumerated powers decisions as “hard,” “substantive” rules).

172. U.S. CONST., amend. X. 173. See New York, 505 U.S. at 155–57. I say “potentially” here because the existence of federal

power alone does not entail the absence of state power—federal and state powers are largely concurrent. However, the existence of federal power can diminish state power in the event of preemptive lawmaking. See Young, Two Federalisms, supra note 1, at 136 (observing that “limits on the Commerce Clause are closely linked to states’ autonomy; those limits, after all, preserve a zone of regulatory authority that Congress may not preempt.”).

174. See, e.g., Bradford R. Clark, The Supremacy Clause as a Constraint on Federal Power, 71 GEO. WASH. L. REV. 91 (2003) (arguing that the Supremacy Clause and separation-of-powers norms effectively limit the federal government’s regulatory reach to the benefit of states); Jackson, supra note 164, at 2228 (“To make political safeguards of federalism work . . . some sense of enforceable limits [on national-government powers] must linger.”).

175. Cf. New York, 505 U.S. at 156 (“In a case . . . involving the division of authority between federal and state governments, the two inquiries are mirror images of each other. If a power is delegated to Congress in the Constitution, the Tenth Amendment expressly disclaims any reservation of that power to the States; if a power is an attribute of state sovereignty reserved by the Tenth Amendment, it is necessarily a power the Constitution has not conferred on Congress.”); see also Pursley, Dormancy, supra note 20, at 514–17 (discussing the relationship between enumeration and reservation under the Tenth Amendment).



enumerated powers question is also necessarily a question about the Tenth Amendment’s reservation of power to the states, then every rule enforcing limitations on the national government’s enumerated powers should be classified as a direct federalism rule.176 That view of enumeration and federalism, however, is disputed and arguably lacks support in the Constitution.177 The alternative is to suppose that some powers are neither delegated to the national government nor reserved to the states—that some things are beyond the capacity of either level of government. This weakens the relationship between enumerated powers doctrines and federalism norms. Uncertainty as to which view is correct makes the Commerce Clause cases difficult to categorize. Other enumerated powers cases are difficult for similar reasons.178 To the extent that doctrines like those established in Erie and Lopez are grounded on constitutional limitations on national power, the relevant limitations might be created by federalism external norms.

Some regard Erie179 as “the most important federalism case of the twentieth century.”180 This view is based at least in part on the volume of discussion of federalism in the Erie opinions; other indirect federalism cases have similarly lengthy discussions of federalism. It is easy to see why these are often treated as federalism cases. My point here is not that cases

176. See, e.g., Young, Two Federalisms, supra note 1, at 32–34 (characterizing Lopez and Morrison as “substantive” federalism rules that “draw[ ] substantive lines between state and federal authority”); Steven G. Calabresi, “A Government of Limited and Enumerated Powers: In Defense of United States v. Lopez, 94 MICH. L. REV. 752 (1995) (arguing that Lopez is “a revolutionary and long overdue revival of the doctrine that the federal government is one of limited and enumerated powers” which raises important questions about “propriety of vigorous judicial review in federalism cases”).

177. See Manning, supra note 7, at 2063–65 (arguing that “the Tenth Amendment cannot provide noncircular justification for the Court’s freestanding federalism. . . . Although . . . [it] contains literally no direction about how to ascertain what powers have been delegated. If notions of federalism can independently limit a delegation to the federal government, the Tenth Amendment of course picks up that limitation; if not, nothing in the amendment supplies the omission.”); see also Pursley, Dormancy, supra note 20, at 512–28 (canvassing arguments for implied national powers).

178. One important example is the Section Five power. Cases interpreting its scope almost inevitably invoke federalism considerations because the Fourteenth Amendment directly restricts state government action that violates the Amendment’s substantive provisions. See Young, Two Federalisms, supra note 1, at 148–49. Like other enumerated powers cases, their basic holdings may be construed as following from several decisive constitutional reasons only one of which is federalism. In City of Boerne v. Flores, 521 U.S. 507, 536 (1997), for example, the Court held that Section Five legislation must be “congruent and proportional” to a pattern of rights violations. Id. at 534–36. This decision has been characterized as directly implementing (1) an interpretation of the text of Section Five; (2) separation-of-powers norms regarding the extent of deference required to congressional enforcement decisions and the nature of the judicial enforcement role under Section Five; and (3) federalism norms precluding overly broad federal regulation of state government conduct. See Young, Two Federalisms, supra note 1, at 148–49.

179. Erie, 304 U.S. at 64. 180. Young, Common Law, supra note 99, at 1657 (2008); see also CHARLES ALAN WRIGHT &

MARY K. KANE, LAW OF FEDERAL COURTS 376 (7th ed. 2011) (“It is impossible to overstate the importance of the Erie decision.”).



like Erie are not “federalism cases” in a broad sense. I argue only that we should distinguish cases in which constitutional federalism norms are the primary reason for decision from those in which they influence the application of other norms. Whether a case falls into one category or the other has little to do with whether it is an important case for understanding courts’ views about federalism.181 That many important federalism cases involve indirect rules underscores the importance of understanding the function of federalism norms where they bear heavily on decisions applying non-federalism norms. Preemption, for example, is arguably a crucial issue for state autonomy—the more state law preempted, the less meaningful policymaking discretion states possess182—but preemption doctrine accounts for federalism only with the presumption against preemption, an indirect rule.183 Conditional spending doctrine raises similar concerns but, after the Court’s rejection of meaningful substantive limitations on conditional federal spending in Dole,184 the spending power is policed only by indirect rules like the Penhurst canon.185

Given the multiple examples of cases in which DF appears to be true and the tensions between DF and the standard model, we should want to know what account of constitutional norms could make DF true. I explore several possible explanations in the next Part.

II. FEDERALISM’S DEFEASIBILITY

We need a conceptual account of constitutional norms that explains the role that federalism norms play in the formulation of indirect federalism rules. Our examples show that federalism norms weigh in the formulation

181. Indeed, some argue that indirect federalism-influenced interpretive canons have become the most important federalism doctrines with the gradual shrinking of categorical federalism-based limitations on government action. See, e.g., Young, Two Federalisms, supra note 1, at 123–27; 130–34; Eskridge & Frickey, supra note 4, at 598.

182. See Egelhoff v. Egelhoff, 532 U.S. 141, 160–61 (2001) (Breyer, J., dissenting) (noting “the practical importance of preserving local independence at retail, i.e., by applying preemption analysis with care, statute by statute, line by line, in order to determine how best to reconcile a federal statute’s language and purpose with federalism’s need to preserve state autonomy”); Young, Two Federalisms, supra note 1, at 131 (“Doctrines limiting federal preemption of state law go straight to the heart of the reasons why we care about federalism in the first place.”); Calvin Massey, Federalism and the Rehnquist Court, 53 HASTINGS L.J. 431, 508 (2002) (arguing “that the failure of the Court to apply preemption doctrine sparingly, and with real attention to both Congress’s intent and the values of federalism, will in the long run prove disastrous to” federalism).

183. See Young, Two Federalisms, supra note 1, at 132 (explaining that “the central aspects of preemption doctrine rely on process mechanisms—in this case, soft rules of statutory construction—to do their work”).

184. South Dakota v. Dole, 483 U.S. 203, 206 (1987) (setting out the basic conditional spending doctrine); see also Eskridge & Frickey, supra note 4 (critiquing Dole).

185. Dole, 483 U.S. at 206; Pennhurst State School & Hospital v. Halderman, 451 U.S. 1 (1981) (holding that spending conditions terms must be set forth unambiguously).



of a variety of doctrines, the primary purposes of which are to implement non-federalism norms. Most commentators skip directly from this observation to normative questions about whether federalism’s observed influence is sufficient or desirable in one context or another.186 Still, the explanation for DF’s truth is not obvious. Without answers, it is difficult to assess the practice of crafting indirect federalism rules. Federalism compatibilists tend to set aside these conceptual questions and assume that courts may legitimately consider a variety of factors, including practicalities, in crafting doctrine.187 Federalism conventionalists make a similar leap, but for different reasons: Federalism is a non-optional part of the constitutional structure; but since there is little text to guide doctrinal formulation, other determinants of doctrine must necessarily be on the table.188 They contend that our focus should be on whether the resulting doctrines display fidelity to the Constitution’s original design.189 Normative federalism scholarship is valuable but nevertheless would benefit from an answer to our basic question. An account of the functioning of federalism norms in indirect cases can ground a framework for assessing the legitimacy of federalism doctrine and, in turn, can provide a sturdier footing for assessing these normative claims.

In assessing potential explanations for the truth of DF, I employ several well-accepted criteria for selecting among competing theories. First, of course, is the criterion of plausibility: We ought to reject an explanation that is facially implausible in the light of what we know of our constitutional system. Second, we should prefer simpler explanations to more complex ones. Third, we should prefer the explanation that makes sense of the largest volume of the phenomena we are trying to explain. And fourth, we should choose an explanation that leaves most of our other well-

186. See, e.g., Young, Two Federalisms, supra note 1, at 123–27 (arguing that clear statement rules are a good strategy for pursuing federalism-related values in some circumstances); Eskridge & Frickey, supra note 4, at 629–44 (canvassing federalism-related interpretive canons and noting conceptual problems briefly before mounting normative critique).

187. See generally sources cited supra note 15 (defending a variety of compatibilist theories of federalism). But see Stuart Minor Benjamin & Ernest A. Young, Tennis with the Net Down: Administrative Federalism Without Congress, 57 DUKE L.J. 2111, 2119 (2008) (highlighting these theorists’ tendency to bypass the question of constitutional legitimacy and criticizing them on that ground).

188. See Young, Federalism Doctrine, supra note 8, at 1736 (arguing that “fidelity to the Constitution requires us to have federalism doctrine”); id. at 1744 (acknowledging that doctrinal formulation often must account for non-interpretive reasons, including “independent moral principles, pragmatic concerns about the workability of particular rules, or institutional issues concerning the Court’s legitimacy”).

189. See, e.g., id. (undertaking this kind of exercise for nearly the corpus of federalism doctrine). See generally Lawrence Lessig, Translating Federalism: United States v. Lopez, 1995 SUP. CT. REV. 125.



settled views about the world intact over one that does not.190 Call these considerations of “plausibility,” “simplicity,” “consilience,” and “conservatism,” respectively; and note that a theory’s success along any one of these dimensions may be offset by deficiency on others.191 Finally, we should reject explanations that are internally incoherent in view of our goal of determining whether indirect federalism doctrines are justifiable. I’ll consider two categories of explanations, one of which can be dealt with briefly.

A. Against the Standard Model

First, there are theories of adjudication that deny one or more central premises of the standard model—one way to quickly dispose of the problem of federalism norms’ seeming failure to follow the standard model of constitutional adjudication is to reject the standard model. And there are several alternative models in the literature. In this section, I survey explanations for DF’s truth predicated on some such accounts and argue that they are ultimately unsatisfying.

Some accounts deny the existence of a straightforward causal relationship between constitutional norms or doctrine and outcomes in constitutional adjudication. These include explanations predicated on realist claims that officials’ actions are better explained or predicted by officials’ personal preferences than by the causal force of legal norms.192 There are several views that build on this kind of claim: “Conceptual rule skepticism,” for example, is the view that “rules previously enacted by legislatures or articulated by courts are not law”; instead, “‘[t]he law is just a prediction of what a court will do’ or ‘[t]he law is just whatever a court says on a particular occasion.’”193 Hart argued that this view cannot explain two common phenomena: judicial mistakes (and our ability to coherently claim that courts have made mistakes);194 and the fact that, in deciding

190. See Brian Leiter, Explaining Theoretical Disagreement, 76 U. CHI. L. REV. 1215, 1239–40 (2009) (citing W.V.O. QUINE & J.S. ULLIAN, THE WEB OF BELIEF (2d ed. 1978); Paul Thagard, The Best Explanation: Criteria for Theory Choice, 75 J. PHIL. 76 (1978)).

191. See Leiter, supra note 190, at 1240. 192. See generally LEE EPSTEIN & JACK KNIGHT, THE CHOICES JUSTICES MAKE 10–11 (1998);

JEFFREY A. SEGAL & HAROLD J. SPAETH, THE SUPREME COURT AND THE ATTITUDINAL MODEL

REVISITED 44–115 (2002). See, e.g., Orley Ashenfelter, Theodore Eisenberg & Stewart J. Schwab, Politics and the Judiciary: The Influence of Judicial Background on Case Outcomes, 24 J. LEGAL

STUD. 257 (1995) (civil rights cases); Frank B. Cross & Emerson H. Tiller, The Three Faces of Federalism: An Empirical Assessment of Supreme Court Federalism Jurisprudence, 73 S. CAL. L. REV. 741 (2000).

193. BRIAN LEITER, Legal Realism and Legal Positivism Reconsidered, in NATURALIZING

JURISPRUDENCE: ESSAYS ON AMERICAN LEGAL REALISM AND NATURALISM IN LEGAL PHILOSOPHY 59, 69 (2007) (citing H.L.A. HART, THE CONCEPT OF LAW 135 (2d ed. 1994)).

194. See LEITER, supra note 193, at 69–70; HART, supra note 193, at 143–44.



cases, courts explicitly ask questions like “What does the relevant statute provide?” rather than “What will I decide the outcome here should be?”195 “Empirical” rule-skeptics take the demonstrated predictive power of officials’ personal preferences to disprove the causal relationship between legal norms and the actions of legal officials.196 They argue that legal norms are (at least sometimes) rationally indeterminate; on this view, even if officials wanted to reach results that were compelled by legal norms, they could not either because the existence of multiple conflicting legal norms makes it impossible to find a single “correct” result or because the interpretability of legal norms makes it impossible to say with certainty that a norm compels a particular result.197

On any of these accounts, DF is easy to explain: If the stated legal reasons for judicial outcomes are not the real reasons for decision, then it does not matter whether federalism norms may legitimately be treated as defeasible, at least not for understanding outcomes in constitutional cases.198 I want to set this kind of explanation aside for several reasons: First, it trivializes the question of DF’s truth, which I want to take seriously. Second, we have too much experience with legal constraint to believe that it simply does not exist—the preference-based account of judicial decision-making is increasingly challenged.199 Indeed, I think we have too much experience with individuals’ paying careful attention to

195. See LEITER, supra note 193, at 70–71; HART, supra note 193, at 102, 143. 196. LEITER, supra note 193, at 73 (noting two skeptical claims: “(1) legal rules are

indeterminate; and, as a result, (2) legal rules do not determine or constrain decisions”); HART, supra note 194, at 138 (noting this view); see, e.g., ROBERT D. COOTER, THE STRATEGIC CONSTITUTION 376 (2000) (noting that economists characterize “an [legal] internal obligation” as “a preference”). But see EPSTEIN & KNIGHT, supra note 192, at 9–10 (assuming that legal constraint exists).

197. See Brian Leiter, Rethinking Legal Realism: Toward a Naturalized Jurisprudence, 76 TEXAS

L. REV. 267, 273–77 (casting American Legal Realism’s core claims as (1) legal norms in appellate cases often are rationally indeterminate; and (2) judges respond primarily to facts rather than legal rules”); see, e.g., BOBBITT, supra note 12; ROBERT L. TSAI, ELOQUENCE AND REASON: CREATING A

FIRST AMENDMENT CULTURE 31–32 (2008); Ian Bartrum, The Constitutional Canon as Argumentative Metonymy, 18 WM. & MARY BILL RTS. J. 327 (2009) (all arguing that constitutional reasoning is to some degree indeterminate such that it bears no direct causal relationship to judicial outcomes).

198. See Leiter, supra note 193, at 64–65 (“For the Realists . . . there is no ‘foundational’ story to be told about the particular decision of a court: legal reasons would justify just as well a contrary result. But if legal rules and reasons cannot rationalize the decisions, then they surely cannot explain them either . . . .”).

199. See Leiter, supra note 190, at 1226–27 (“[T]he main reason the legal system of a modern society does not collapse under the weight of disputes is precisely that most cases . . . never go any further than the lawyer’s office” an very few proceed to appellate review where some indeterminacy is possible); Frederick Schauer, Easy Cases, 58 S. CAL. L. REV. 399, 429–30 (1985) (both arguing that experience demonstrates that officials most often agree on the law and act accordingly). Skepticism exists, but most positive legal scholarship is concerned with prediction rather than metaphysical or psychological truths. See Richard H. Fallon, Jr., Constitutional Constraints, 97 CAL. L. REV. 975, 977–78, 981 (2009) (characterizing much law and economics literature as focused on modeling behavior according to non-legal factors); John Ferejohn, Positive Theory and the Internal View of Law, 10 U. PA. J. CONST. L. 273, 275 (arguing that positive theory can account for legal constraints).



legal reasons to believe, without more empirical evidence, that they do not matter to outcomes of cases.200 Hart rejected global rule skepticism, arguing that “though every rule may be doubtful at some points, it is indeed a necessary condition of a legal system existing, that not every rule is open to doubt on all points.”201

Indeterminacy theories cannot account for the bulk of our everyday experience with the legal system202—which is the experience of “massive and pervasive agreement about the law throughout the system. It is precisely because just about everyone agrees about the law that lawyers can tell most prospective clients who wander through the door that they have no claim and should go home . . . it is precisely because just about everyone agrees about the law that most cases are not appealed; and so on.”203 Doctrine makes a difference; it crystallizes the legal requirements that form the basis of widespread agreement about law’s constraints. We therefore have a reason to want to understand and rationalize doctrine and resolve its curiosities where possible.

Even if we do not know what reasons figure in the best causal account of outcomes in constitutional adjudication, DF is a puzzle for the internal coherence of the system of legal reasons. If justifications involving legal reasons are ultimately inconsistent with real judicial motivations; legal reasons still are phenomena in the world worth understanding. The American Legal Realists did not dismiss the practical importance of understanding legal doctrine—they agreed that one must always provide courts with “a technical ladder” of legal reasons to justify a result; though one must also “on the facts. . .persuade the court” to rule in one’s favor.204

200. See HART, supra note 193, at 141 (asserting that “it is surely evident that for the most part decisions . . . are reached either by genuine effort to conform to rules consciously taken as guiding standards of decision or, if intuitively reached, are justified by rules which the judge was antecedently disposed to observe and whose relevance to the case in hand would generally be acknowledged”). Leiter rightly criticizes Hart’s argument as “no argument at all. He simply denies what the Realist affirms, but gives no reason for the denial other than his armchair confidence in the correctness of his own view.” LEITER, supra note 193, at 78. Empirical evidence that judicial decisions are caused by non-legal reasons must be countered by empirical evidence documenting the causal force of legal norms; and that research is only beginning. See, e.g., Fallon, supra note 199, at 995–1002 (framing the inquiry as about “normative” and “external” constitutional constraints).

201. HART, supra note 193, at 152; see id. at 141–53 (critiquing “rule skepticism” by arguing, first, that the possibility of unconstrained decisionmaking does not make it inevitable, and, second, with respect to judicial discretion in cases with indeterminate legal norms like ambiguous statutes, that discretion is intentionally conferred by secondary rules of jurisdiction: “courts have jurisdiction to settle [disputes] by choosing between the alternatives which [a] statute leaves open, even if they prefer to disguise this choice as a discovery”).

202. Cf. Leiter, supra note 190, at 1228 (“[L]egal positivism . . . explains the pervasive phenomenon of legal agreement.”).

203. Id. at 1227. 204. LEITER, supra note 193, at 77 (quoting KARL N. LLEWELLYN, THE BRAMBLE BUSH 76

(1930)).



One possibility, of course, is that there is no coherent explanation that justifies indirect federalism rules; if that is true, it is worth knowing because it might prove an important datum for those who think the system of constitutional reasons underdetermines outcomes. Even if critics are right to charge that the normative agenda of legal process theorists205—a call for constitutional adjudication according to “neutral principles” and with rational legal explanations for constitutional holdings206—merely provides stage dressing for preference-driven judicial decisionmaking, we still must operate in a system in which courts are empowered to act authoritatively.207 We might tolerate hidden rational indeterminacy in the system of legal reason and the possibility that judicial decisionmaking is results-oriented if we accept an account of legitimacy predicated on the public presentation of coherent legal reasons. But rational incoherence in the system of legal reasons would suggest illegitimacy too powerfully to ignore, undermining even this modest criterion of judicial legitimacy. DF’s truth, then, is important even if we accept a realist account; and in trying to explain it we will profit from assuming arguendo that constitutional doctrine does bear a causal relationship to judicial decisions.

Other explanations for DF’s apparent truth may be predicated on denying the centrality of “norms” in the standard model. For example, we might posit an additional category of constitutional inputs—call them “principles”—that carry weight in adjudication distinct from that accorded to norms in the standard model.208 To avoid confusion, note that commentators frequently invoke “principles” loosely to refer to abstract constitutional norms derived primarily by inference form the constitutional

205. See generally HENRY M. HART, JR. & ARTHUR M. SACKS, THE LEGAL PROCESS: BASIC

PROBLEMS IN THE MAKING AND APPLICATION OF LAW (William N. Eskridge, Jr. & Philip P. Frickey, eds. 2001); HART & WECHSLER, supra note 157.

206. See generally Herbert Wechsler, Toward Neutral Principles of Constitutional Law, 73 HARV. L. REV. 1 (1959) (setting out the Legal Process agenda, urging constitutional adjudication according to “neutral principles” to avoid delegitimizing the courts). Wechsler’s piece is infamous for arguing that Brown v. Board of Education, 347 U.S. 483 (1954), was either wrongly decided or wrong in its articulation of reasons for decision. Wechsler, supra, at 32–33; see Dan M. Kahan, The Supreme Court 2010 Term—Foreword: Neutral Principles, Motivated Cognition, and Some Problems for Constitutional Law, 125 HARV. L. REV. 1, 10–11 (2011) (“What . . . shocked [Wechsler’s readers] . . . was his conclusion that the Supreme Court’s decision five years earlier in Brown v. Board of Education did not pass his ‘neutral principles’ test.”).

207. See, e.g., Jenny S. Martinez, Process and Substance in the “War on Terror,” 108 COLUM. L. REV. 1013, 1025 (2008); William N. Eskridge, Jr., Metaprocedure, 98 YALE L.J. 945, 964 (1989) (reviewing ROBERT M. COVER, OWEN M. FISS & JUDITH RESNIK, PROCEDURE (1988)) (both noting the argument that legal process’s focus on process may “submerge critical debate over substantive justice”).

208. Dworkin is perhaps the only theorist to advance a fully rendered view of this kind. See, e.g., RONALD DWORKIN, LAW’S EMPIRE 225–75 (1986) (describing a process of “integrity”-seeking adjudication in which judges attempt to discern broad principles that best “fit” the past practice of the legal system to apply in each case).



text.209 Think here of arguments that proper inferences from the constitutional text require, for example, democratic governing institutions and processes, equal treatment and minimally decent living conditions for everyone, and so forth.210 So far, I have been distinguishing legal norms by their position on our normative hierarchy rather than their level of abstraction.211 Constitutional norms may be broad and abstract—to fit the definition of “constitutional norm” on the standard model, they need only be derivable from the Constitution and carry the superordinate normative status attributed to constitutional norms.212 That superordinate status is exactly what federalism norms appear to lack in the indirect cases; thus if these principles are to account for indirect cases—that is, if we can explain federalism’s role in those cases by claiming they involve federalism principles rather than federalism norms—then principles must have less normative force than constitutional norms. I do not think that is what most casual invocations of constitutional principles mean to suggest—some proponents of constitutional principles that are somehow distinct from norms maintain that all “constitutional concerns must be immediate and nonnegotiable to be effective”—in other words, these constitutional principles need to have the same normative priority as constitutional norms to do the work ascribed to them.213

Consider a more deliberate and nuanced use of the concept of a constitutional “principle.” Dworkin describes constitutional adjudication as a process of “constructive interpretation” whereby courts determine in each case which abstract moral principle casts relevant past legal practice in the best moral light and then decide according to that principle.214 Dworkin’s principles, adduced by interpretation of past legal practice and by reference to paradigmatically moral principles of “justice, fairness, and procedural due process,”215 appear to be the most normatively powerful inputs in constitutional adjudication on Dworkin’s view, trumping even conventional

209. See, e.g., Young, Federalism Doctrine, supra note 8, at 1746 (“It may help to begin by adding a third category of constitutional ‘law’ . . . [which] would include fundamental structural principles . . . bedrock ideas undergirding the textual provisions in the document and tying them together into a coherent structure.” (emphasis added)). I do not think that Professor Young’s “principles” are distinct from what I am calling “norms.”

210. See, e.g., sources cited supra note 42; DWORKIN, supra note 208, at 382. 211. See supra notes 26–40 and accompanying text. 212. Cf. Fallon, supra note 28, at 62 (noting that “some constitutional norms may be too vague to

serve directly as effective rules of law” such that “a perfect correspondence could not, even in principle, exist between the meaning of constitutional norms and the doctrinal tests by which those norms are implemented (emphasis added)).

213. SAGER, supra note 42, at 154. 214. See DWORKIN, supra note 208, at 225 (“According to law as integrity, propositions of law

are true if they figure in or follow from the principles of justice, fairness, and procedural due process that provide the best constructive interpretation of the community’s legal practice.”).

215. DWORKIN, supra note 208, at 225.



interpretations of constitutional norms that require results different from those most consistent with principle.216 Dworkin’s view has drawn a broad array of criticism;217 most important here, however, is the observation that his principles do not straightforwardly explain federalism’s apparent function in indirect cases. We need something with less normative priority than conventional constitutional norms (never mind that Dworkin’s principles are moral principles, and it is hard to imagine a federalism principle fairly characterized as a requirement of justice).218 Nevertheless, Dworkin’s principles may constitute a distinct form of normative input not accounted for in the standard model and, thus, suggest a use of the concept of “principles” to explain DF’s apparent truth.219 We would need sub-constitutional principles, including some that have to do with federalism, carrying normative priority inferior to that of standard constitutional norms but that nevertheless are properly considered weighty legal reasons in judicial decisionmaking.

The problem with a principle-based account of federalism’s defeasibility is that we lack an accepted and well-developed ontology of constitutional principles in our conventional account of constitutional practice or in the constitutional theory literature. Such an ontology is necessary to explain both the distinctive normative status of these principles and how they relate causally to judicial outcomes. Remember

216. See DWORKIN, supra note 208, at 243 (arguing that “the law is structured by a coherent set of principles” and that courts should “enforce these in the fresh cases that come before them, so that each person’s situation is fair and just according to the same standards.”).

217. Dworkin’s critics, as Professor Schauer has said, “are legion.” Frederick Schauer, (Re)taking Hart, 119 HARV. L. REV. 852, 873 (2006) (reviewing NICOLA LACEY, A LIFE OF H.L.A. HART: THE

NIGHTMARE AND THE NOBLE DREAM (2004)). The following is just a sampling. See, e.g., SCOTT SHAPIRO, LEGALITY 307–30 (2011) (arguing that Dworkin’s “constructive interpretation” fails to account for limited human capacity); Ian P. Farrell, On the Value of Jurisprudence, 90 TEX. L. REV. 187, 216–23 (2011) (reviewing SHAPIRO, supra) (canvassing newer critiques of Dworkin); John Gardner, Law’s Aims in Law’s Empire, in EXPLORING LAW’S EMPIRE: THE JURISPRUDENCE OF RONALD

DWORKIN 207 (Scott Hershovitz ed., 2006); Joseph Raz, Dworkin: A New Link in the Chain, 74 CALIF. L. REV. 1103 (1986) (reviewing RONALD DWORKIN, A MATTER OF PRINCIPLE (1985)); JOSEPH RAZ, BETWEEN AUTHORITY AND INTERPRETATION: ON THE THEORY OF LAW AND PRACTICAL REASON 79–85 (2009); H.L.A. HART, Between Utility and Rights, in ESSAYS IN JURISPRUDENCE AND PHILOSOPHY 198, 208–21 (1983); RICHARD A. POSNER, THE PROBLEMS OF JURISPRUDENCE (1990) (denying Dworkin’s “right-answer thesis” that courts can reach objectively correct answers in hard cases).

218. I am taking from Dworkin an isolated point. He distinguishes standard legal rules from principles, which are not validated by a system’s conventionally created criteria of legal validity. That distinction could hold and could do theoretical work even if both sets of adjudicatory inputs can have “rule-like” and “principle-like” content. See Schauer, supra note 217, at 873–74 & nn.69–71.

219. This point is contested—some argue that Dworkin’s distinction between rules and principles fails to distinguish anything outside the category of “legal rules” or, as I have phrased it here, “legal norms,” properly conceived. See, e.g., FREDERICK SCHAUER, PLAYING BY THE RULES: A

PHILOSOPHICAL EXAMINATION OF RULE-BASED DECISION-MAKING IN LAW AND IN LIFE 12-14 (1991); Joseph Raz, Legal Principles and the Limits of Law, in RONALD DWORKIN AND CONTEMPORARY

JURISPRUDENCE, supra note 217, at 73, 82 (both arguing that Dworkin’s distinction between rules and principles is false).



that the concept of a constitutional “norm” as it figures in everyday use designates a constitutional requirement, preclusion, or permission simpliciter, which by definition carries the superordinate normative priority of constitutional norms; the standard model recognizes no intermediate level of normative priority between that of constitutional norms and that of ordinary legislative, administrative or common law norms.220 Before we can resolve DF’s truth with these principles, then, we confront significant questions about these principles themselves: What, exactly, distinguishes constitutional principles from constitutional norms? Even with a criterion of demarcation, how should we determine which kind of requirement—principle or norm—a particular provision of the Constitution generates? And, most importantly, we still need to know how constitutional principles have normative priority inferior to that of conventional constitutional norms. Positing intermediately normative federalism principles as an explanation for indirect federalism cases appears to reproduce the problem.

My claim here is modest. Periodic invocation of constitutional principles suggests it is plausible that something other than constitutional norms may function as an input in constitutional adjudication. But the principle-based explanation does not display simplicity—to work, it requires constructing a complex conceptual account to support the existence and functioning of legal principles with characteristics that would explain DF. Consilience also may be a problem: It is not clear what these principles would explain beyond indirect federalism cases and similar indirect applications of other structural norms. Finally, the principles solution seems to lag on the conservatism criterion, at least insofar as the standard model centers on norms as the primary constitutional input in adjudication.

Similar to the principles account in their tendency to reproduce the DF question are explanations that depend on particularized accounts of constitutional adjudication to provide a foundation for federalism norms’ defeasibility in indirect cases. On Dean Ely’s view, for example, constitutional adjudication is justified by the “meta-neutral principle”221 of “representation reinforcement”: courts should intervene “only when the. . .political market[] is systematically malfunctioning”—where the process of representative democracy, which ordinarily can be trusted to safeguard individual rights and systemic stability, becomes unreliable in that

220. See supra notes 26–40 and accompanying text. We might posit principles with ordinary, not intermediate, normative priority; but that leaves us with a tension between our conventional views about federalism as constitutional in stature and our account of its place in the hierarchy of legal authority.

221. See Kahan, supra note 206, at 13 (employing this term to describe the targets of certain theories of constitutional adjudication, including those of Herbert Wechsler, Alexander Bickel, John Hart Ely, and Robert Bork, among others).



function.222 Ely juxtaposes this account with “interpretivism” and a “value-protecting approach,”223 which map roughly onto two sides of the debate about interpretive theory that occupies much contemporary constitutional literature—originalism and “living constitutionalism.”224 By denying the centrality of interpretive theories, Ely offers an account of constitutional adjudication that focuses on the legitimacy of judicial intervention into the political process rather than on the substantive correctness of individual outcomes.225 Professor Young’s proposed “Democracy and Distrust for federalism,” modeled on Ely’s view, thus might provide a different approach to our question.226

Echoing Ely, Young argues that constitutional adjudication should shore up process protections with a focus on institutional protections for the federal structure.227 Young’s view thus relies on the supposed political and procedural safeguards of federalism;228 but it is moderate in that it sanctions more frequent direct judicial enforcement of constitutional federalism norms than does strict process federalism theory.229 Young argues that courts justifiably intervene with doctrine that makes “compensating adjustments” when the balance of power between the federal and state governments shifts too dramatically in one direction or another—when the

222. ELY, supra note 42, at 103. Ely explains that political malfunction is evident where the process is undeserving of trust, when (1) the ins are choking off the channels of political change to ensure that they will stay in and the outs will stay out, or (2) though no one is actually denied a voice or a vote, representatives beholden to an effective majority are systematically disadvantaging some minority out of simple hostility or a prejudiced refusal to recognize commonalities of interest, and thereby denying that minority the protection afforded other groups by a representative system. Id.

223. See id. at 87–88 & n.*. 224. We might say Ely’s “interpretivists” have become today’s originalists and textualists, and

that his “value-protectors” are today’s living constitutionalists. But the correctness of this sub-characterization is not important here.

225. See John Hart Ely, Toward a Representation-Reinforcing Mode of Judicial Review, 37 MD. L. REV. 451, 485–87 (1978) (noting his preference for judicial emphasis on process rather than on substantive outcomes). Ely’s view has, of course, encountered substantial criticism. See, e.g., Laurence H. Tribe, The Puzzling Persistence of Process-Based Constitutional Theories, 89 YALE L.J. 1063, 1067–72 82 (1980) (canvassing critical arguments).

226. Young, Two Federalisms, supra note 1, at 16. 227. Id. 228. See Young, Federalism Doctrine, supra note 8, at 1753 (“[C]ounting on the Court to play a

primary role in [the preservation of the federal balance] seems unrealistic, and the most promising strategies for maintaining some sense of balance in our system will need to pursue action across a variety of legal, political, and private institutions.”) (emphasis in original).

229. See generally Wechsler, supra note 50 (giving the canonical statement of the political safeguards view); Garcia v. San Antonio Metro. Transit Auth., 469 U.S. 528, 552 (1985) (federalism values “are more properly protected by procedural safeguards inherent in the structure of the federal system than by judicially created limitations on federal power”). For the strict political safeguards view, which suggests that federalism should be a non-justiciable political question outside the scope of judicial review, see, for example, JESSE H. CHOPER, JUDICIAL REVIEW AND THE NATIONAL POLITICAL

PROCESS: A FUNCTIONAL RECONSIDERATION OF THE ROLE OF THE SUPREME COURT (1980).



basic character or stability of the system is threatened.230 This view may explain why courts have considered federalism in the variety of cases in which we observe it influencing the decision. The federalism implications of judicial implementation of other norms can suggest that the federal structure is threatened even absent a direct violation of federalism norms.231 If courts must intervene whenever federalism is threatened, they may need to be opportunistic in selecting cases in which to make compensating adjustments.232 Indeed, Young points to some indirect federalism cases as having created centrally important federalism doctrine. Young’s account, therefore, may explain why we observe indirect federalism rules; but it also, again, highlights the absence of an account of how constitutional norms may function as defeasible reasons for decision.233

Young’s theory tells us little about the actual content, and, more importantly here, the conceptual status of federalism norms. That the need for compensating adjustments to preserve a balance of federal and state power justifies judicial intervention, without more, does not entail anything about the nature of that judicial intervention except, perhaps, what Young calls its “direction”—toward centralization or decentralization depending on the kind of imbalance that prompts intervention.234 Young emphasizes that the success of his account requires no position on the content of federalism norms.235 This account does not explain how even an abstract, simple constitutional federalism norm requiring rough balance may function defeasibly sometimes but with superordinate normativity at others. It reproduces our question: It gives a “when” and something about the “how” of judicial enforcement of federalism norms (namely, courts upon

230. See Young, Federalism Doctrine, supra note 8, at 1756–58 (listing historical examples of imbalances requiring adjustment); id. at 1775 (arguing that compensating doctrines are justified “even where such doctrines are not derivable directly from the text and history of the Constitution”).

231. Young suggests something like DF in arguing that courts should look at constitutional doctrine more “holistically” to assess whether it maintains the balance of federalism. Cf. Young, Federalism Doctrine, supra note 13, at 1756–57 (arguing that “‘a court cannot craft an optimal doctrinal rule without considering the interactive effects between it and other rules’” (quoting Evan H. Caminker, Context and Complementarity Within Federalism Doctrine, 22 HARV. J. L. & PUB. POL’Y 161, 161 (1998))).

232. See supra note 231. Cf. Young, Federalism Doctrine, supra note 8, at 1758 (“Any institution charged with helping to maintain the federal balance must be alert to the need, at different times, to throw its weight onto one or the other side of the scale.”).

233. See, e.g., Young, Federalism Doctrine, supra note 8, at 1759–60 (discussing Lopez); Young, Two Federalisms, supra note 1, at 8–9 (counting as instances of federalism doctrine a variety of indirect cases).

234. See Young, Federalism Doctrine, supra note 8, at 1756–58 (arguing that courts must determine the “direction” of adjustment and and that different directions may be appropriate at different times).

235. See id. at 1804 (“I want to reject the notion that courts must identify a particular point of optimal or correct equilibrium between national and state power in order to make federalism doctrine.”).



identifying process problems should formulate doctrine that accomplishes compensating adjustments); but it does not tell us what we need to know about the “what”—the conceptual status of the norms themselves.236

Perhaps a better way to approach the question of DF’s apparent truth is to start within the framework of the standard model of constitutional adjudication.

B. Instrumental Reasons and Underenforcement

The second category of potential explanations that I want to consider casts federalism’s defeasibility in the indirect cases as resulting from some feature of the relationship between constitutional norms and the rules, tests, and standards courts formulate to implement those norms. Assuming arguendo that the legal reasons provided in judicial opinions do, in fact, stand in a causal relationship with adjudicatory outcomes; we should construct an account of the kind of legal reason that federalism might be when it functions as a strong yet defeasible reason in doctrinal formulation. There seem to be two main possibilities: First, the doctrines formulated and applied in indirect cases might be instances of judicial underenforcement of federalism norms that would, but for that underenforcement, function decisively. Second, these cases may involve circumstances in which federalism norms just have non-decisive normative status. Put simply, rather than think of the indirect rules as concerned with enforcing federalism norms through unfamiliar or obscure means; we may think of them as concerned with enforcing other constitutional norms through a doctrine-making process that incorporates non-decisive consideration of federalism. The question is whether courts can legitimately do that—I argue that they can, provided that we may legitimately posit a new category of reasons—”quasi-constitutional” reasons—to be added to the standard model of constitutional adjudication.

To better understand the role federalism norms plays in the formulation of indirect federalism rules, it is helpful to emphasize the distinction between the Constitution’s basic norms and the rules, tests and standards courts use to enforce those norms.237 Constitutional theorists increasingly recognize an important distinction between categories of constitutional doctrine expressed by the “two-output thesis,” viz.: the “claim that there exists a conceptual distinction between two sorts of judicial work product each of which is integral to the functioning of constitutional adjudication,

236. He suggests that they may be underenforced and cautions against judicial involvement unless there is a need for a compensating adjustment; we will return to the point in the next section. See infra notes 255–266 and accompanying text.

237. See supra notes 28–43 and accompanying text.



namely judge-interpreted constitutional meaning and judge-crafted tests bearing an instrumental relationship to that meaning.”238

I want to avoid debates about interpretive method; so call statements of judge-interpreted constitutional meaning constitutional “operative propositions,” following Professor Berman’s interpretively inert terms.239 Operative propositions are doctrine; they are court-formulated statements of interpreted constitutional meaning.240 This is distinct from “correct” or “incontestable” meaning—courts can make interpretive mistakes and the Constitution does not mean whatever a court says that it means.241 The mediating adjudicative rules, tests, and standards are influenced in part by the requirement that they implement the operative proposition—that they “fit” the norm to some degree. Thus, we can say that one category of reasons at work in doctrinal formulation is interpretive reasons. Distinct from operative propositions are the judge-crafted tests and standards of constitutional doctrine—the “decision rules” by which courts determine whether conduct falls within the meaning of a constitutional prohibition or permission, and are distinct from the constitutional operative propositions themselves.242 Decision rules implement operative propositions by instructing courts how to determine whether constitutional norms are satisfied in concrete cases.243

In addition to the interpretive reasons shaping doctrine, courts formulate decision rules based in large part on instrumental reasons that relate to the effectiveness of adjudication.244 Minimizing the risk of adjudicatory error is, perhaps, the most significant consideration in this

238. See Mitchell N. Berman, Aspirational Rights and the Two-Output Thesis, 119 HARV. L. REV. F. 220, 220 (2006) [hereinafter Berman, Rights]; see also sources cited supra note 28.

239. Berman, Rules, supra note 6, at 57–58 & n.192. 240. See Berman, Rights, supra note 238; Mitchell N. Berman, Guillen and Gullibility: Piercing

the Surface of Commerce Clause Doctrine, 89 IOWA L. REV. 1487, 1519–20 (2004) [hereinafter Berman, Guillen].).

241. See Berman, Guillen, supra note 240, at 1519; Roosevelt, supra note 28, at 1661. 242. Berman, Rules, supra note 6, at 57–58 (calling these rules, tests, and standards

“constitutional decision rules.”). 243. See Fallon, supra note 28, at 61–66 (describing constitutional “implementation”). See

generally RICHARD H. FALLON, JR., IMPLEMENTING THE CONSTITUTION 38–44 (2001). 244. There is debate about whether operative propositions are or should be formulated without

reference to instrumental considerations. Berman, supra note 6, at 45–46, 50; see, e.g., Levinson, supra note 28, at 873; Strauss, supra note 28, at 207 (arguing that instrumental concerns are considered at every stage of adjudication). Here I need only the distinction between operative propositions and decision rules, which requires no commitment to a theory of interpretation that excludes pragmatic concerns as relevant. The two-output thesis, understood in this modest sense, is widely accepted. Berman, Rules, supra note 6 at 12–15. Even pragmatists “are not committed to denying the utility of distinguishing between constitutional meaning and judge-crafted doctrine. We need such a distinction to be able to criticize judicial doctrine as based on a misunderstanding of constitutional meaning, and pragmatists need not think such criticisms unintelligible or pointless.” Berman, Rights, supra note 238, at 224; Roosevelt, supra note 28, at 1661.



process;245 but institutional competence concerns are also relevant—for example, deferential decision rules may be adopted to offset the risks of error inherent in deciding questions beyond judicial competence.246 Other instrumental concerns shape decision rules in other contexts.247 The harmfulness of the potential constitutional violation is also relevant: If violations are harmful enough to outweigh the costs of erroneous invalidation of permissible actions, a stricter rule may be desirable for its capacity to “invalidate many unconstitutional laws and very few legitimate ones.”248 Decision rules often leverage characteristics common to a cluster of cases—proxies—that reliably signal constitutional violations; designing rules to be triggered by these proxies is a way to reduce the risk of adjudicatory error. In some instances, the balance of instrumental reasons may support rules prohibiting more or less conduct than violates the underlying constitutional operative propositions. Underenforcement occurs where decision rules uphold some actions that violate the operative proposition—e.g., rational basis review, which “upholds [some] violations but strikes down almost no valid acts.”249 Overenforcement occurs where decision rules invalidate more actions than actually violate the operative proposition; strict scrutiny, for example, “strikes down [some] valid laws but upholds almost no violations.”250

The distinction between operative propositions and decision rules suggests a possible explanation for indirect federalism rules—those rules may operate as they do because they are designed to underenforce federalism norms. Several instrumental reasons support underenforcement in the federalism context: First, judicial attempts to directly limit national action on federalism grounds have resulted in errors and unworkable rules like the now-abandoned National League of Cities doctrine.251 An approach

245. See Berman, Guillen, supra note 240, at 1522 (stressing error-cost considerations in decision-rule formulation); Roosevelt, supra note 28, at 1662–63 (same).

246. Roosevelt, supra note 241, at 1659–60; Berman, Guillen, supra note 240, at 1523 (same); see generally Richard J. Fallon, Judicial Manageable Standards and Constitutional Meaning, 119 HARV. L. REV. 1274 (2006) (discussing the influence of institutional competence on doctrine).

247. For example, concerns about fair representation in the legislative process might commend a non-deferential decision rule. See Roosevelt, supra note 28, at 1664–65. See generally United States v. Carolene Prods. Co., 304 U.S. 144, 152 n.4 (1938); ELY, supra note 42, at 135–37. A need to engage burdensome inquiries into legislative purposes might call for a rule directing courts to presume the purpose’s absence or presence under certain circumstances. Berman, Rules, supra note 6, at 67; Roosevelt, supra note 28, at 1665–66. If other actors need judicial guidance, then clear, formalistic rules are preferable to multi-factored standards. Berman, Guillen, supra note 240, at 1522; Roosevelt, supra note 28, at 1666–67.

248. Roosevelt, supra note 241, at 1663–64; see Berman, Rules, supra note 6, at 75–76, 80 n.253. 249. Roosevelt, supra note 241, at 1661; see also Berman, Rules, supra note 6, at 81; Sager,

supra note 10 (discussing underenforcement). 250. See Roosevelt, supra note 241, at 1661. 251. See supra notes 61–65 and accompanying text; Young, Two Federalisms, supra note 1, at

126 (noting the Court’s “long and discouraging history of failures to formulate” substantive federalism



to federalism centered on statutory interpretation rules might be preferable for avoiding errors. Second, aside from a few clear cases, courts face comparative institutional capacity deficits in determining whether actions undermine the balance or stability of the federal structure.252 That, too, increases the risk of adjudicatory error when courts attempt to directly enforce federalism norms.253

Return to federalism-influenced interpretive canons and remember that, although they are indirect federalism rules, they also operate in contexts that scholars consider crucial for federalism today.254 Some of these canons thus seem to be squarely “about” federalism.255 If we assume an operative proposition that tracks our hypothetical “rough balance” federalism norm, then, relying on the protections for federalism supposedly built into federal lawmaking processes, courts might conclude that deference to coordinate branches is an optimal enforcement strategy in the light of courts’ perceived institutional capacity deficit on federalism.256 Accordingly, rather than adopt rules invalidating actions that the courts conclude are disruptive; courts might choose to underenforce the norm with decision rules—like these interpretive canons—that defer to congressional decision-making about federalism.257 If the institutional capacity claim is correct, then courts might reasonably expect better enforcement of federalism norms from congressional deliberation. Federalism-influenced interpretive canons are designed to force Congress to engage federalism issues directly and enact unambiguous language to avoid narrowing judicial interpretations.258 If

limitations); Garcia v. San Antonio Metro. Transit Auth., 469 U.S. 528 (1985) (rejecting National League of Cities as “unworkable”).

252. See generally Young, Federalism Doctrine, supra note 8, at 1816–20, 1831–44. 253. Both institutional competence and adjudicatory error risks are paradigmatic instrumental

considerations that shape decision rules on the two-output model. See Berman, Rules, supra note 6. 254. See supra notes 127–153 and accompanying text; Eskridge & Frickey, supra note 4. 255. See Young, Two Federalisms, supra note 1, at 16–23 (characterizing federalism-related

canons as of roughly equal importance to direct federalism doctrines). 256. See id. at 126 (arguing that federalism-related canons “function well in areas in which the

relevant constitutional principles are designed primarily to be self-enforcing, through . . . political processes”).

257. This underenforcement account of federalism-influenced canons seems like one that Young would accept. Cf. Young, Resistance Norms, supra 11, at 1585 (arguing that “‘clear statement rules’ . . . are the best way—and perhaps the only way—of giving voice to constitutional norms that are . . . generally left underenforced by more conventional types of doctrines.” (emphasis added)); Benjamin & Young, supra note 187, at 2148 (noting that Young has “defended clear statement requirements as ‘resistance norms’ that can compensate for the underenforcement of federalism and other structural values”).

258. See Young, Two Federalisms, supra note 1, at 126 (arguing that clear statement rules “increas[e] the political costs of . . . government action” by increasing drafting hurdles and mobilizing opposition interests). This approach might be doubly underenforcing: Federalism-influenced canons do not apply universally across statutory interpretation, much potentially disruptive action is not subject to these rules. The Court might choose to construct rules for areas with good proxies for federalism problems—preemption, spending conditions, sovereign immunity, etc.—to avoid adjudicatory error.



Congress enacts clear language, these rules are satisfied and the statute is construed according to its plain meaning.

One problem with an underenforcement account of the federalism canons is that it only works if courts accept federalism norms that allow for some judicial enforcement (like our hypothetical norms)—after all, the application of interpretive canons is different from holding federalism issues entirely nonjusticiable. But the content of federalism norms, again, is contested: If they require nothing more than observance of mandatory lawmaking processes, then federalism canons constitute unnecessary judicial intervention—overenforcement, not underenforcement.259 Moreover, although the Court embraced process federalism generally in Garcia,260 most process federalism theories are inconsistent with the hard federalism doctrines courts apply in the anticommandeering and sovereign immunity contexts.261 Those rules make it unclear what federalism norms courts accept.262 The underenforcement account seems to require more consensus on the content of federalism norms than exists among courts.

Indeed, it is not clear how these rules underenforce anything at all if they do not create the possibility of invalidation. Underenforcement, although invalidating fewer than all violations, is nevertheless direct enforcement of a constitutional norm; and constitutional violations identified by underenforcing rules are invalidated with the standard finality. Now, once we acknowledge the possibility of strategic underenforcement, there is no conceptual obstacle to courts formulating decision rules with no invalidation function. A corollary, however, is that underenforced norms remain binding on non-judicial actors to their full conceptual limits, sweeping into the category of unconstitutional actions all those that the judicial rule does not catch—non-judicial actors remain obligated to avoid constitutionally invalid conduct. This is consistent with our standard account of the status of constitutional norms: Regardless of the courts’ enforcement approach, they are fully normative; nothing can overcome the problem if a constitutional norm is violated, even where the violation will go without judicial remedy. If federalism norms require anything more than the strict-process theory claims, then we should be able to identify violations of those norms even if judicial doctrine does not. But federalism-influenced canons suggest no criterion of constitutional invalidity, in their

259. On the strict process federalism view, see generally sources cited supra note 50. 260. Garcia v. San Antonio Metro. Transit Auth., 469 U.S. 528, 550 (1985). 261. See supra notes 52–67; Young, Two Federalisms, supra note 1, at 38 (characterizing “the

anticommandeering principle as a hard limit on national power: Congress cannot overcome that principle by clearly stating its intent to do so, and the doctrine as stated does not yield before sufficiently important federal interests”).

262. See, e.g., Lawrence H. Tribe, The Puzzling Persistence of Process-Based Constitutional Theories, 89 YALE L.J. 1063 (1980) (canvassing criticism of the process theory).



application or otherwise. Adherence to the rules’ process requirements is sufficient for validity, regardless of the statute’s actual effect on federalism. These rules leave nothing else that is “binding” on non-judicial actors; their animating federalism concerns are fully overridable by ordinary legislation. This creates a legitimacy problem if these canons do, in fact, leave all violations of underlying federalism norm without remedy. In any event, federalism canons do not map well onto the accepted description of underenforcing relationships between norms and decision rules. If they are examples of judicial enforcement or underenforcement of federalism norms, we have the same basic puzzle: Federalism norms are treated as overridable, contrary to the standard model.

Variations in the form and content of indirect federalism rules suggest additional general problems with an underenforcement-based explanation. First, underenforcement seems particularly inapt for explaining courts’ consideration of federalism in cases that are squarely about other constitutional norms—in constitutional rights cases, for example. There, it just seems incorrect to say that the court is enforcing federalism norms at all.263 An underenforcement account is particularly difficult to construct for such cases. For underenforcement to explain federalism’s influence on, say, due process doctrine, we need to know what that doctrine would look like if federalism norms were fully enforced. Absent that information, we cannot specify the degree of underenforcement or the instrumental concerns on which it is based. Moreover, courts in indirect federalism cases often seem to consider federalism out of concern that their own process of doctrinal formulation will adversely effect constitutional the system. For example, federalism would provide a reason to reject a new rule increasing judicial scrutiny for a category of discrimination cases currently adjudicated on the rational basis standard insofar as such a shift threatens to invalidate a greater valume of state action. Here, the doctrine itself creates the potential federalism problem. But if federalism norms restricting judicial action account for the consideration of federalism in some indirect cases; underenforcement cannot explain federalism’s influence on the resulting doctrine. Courts must comply fully with federalism requirements binding on the judiciary; strategic underenforcement of those norms would straightforwardly violate them.264 Our question then becomes more

263. Courts may be balancing conflicting constitutional norms in some of these cases, but we lack a full account of the criteria by which courts should decide which of two or more equally superordinate norms should control. That, however, is not underenforcement—a rule prioritizing conflicting constitutional norms is closer to what we get with the variable normativity hypothesis. See infra Part II.C.2.

264. Some indirect federalism rules might be explained as direct enforcement of federalism norms that require this kind of judicial deliberation about doctrine’s federalism impacts. But it is not clear that federalism norms require that and, even if it were, this would not explain all indirect rules.



pressing: If courts in indirect cases are considering federalism requirements for judicial action, why do federalism norms appear so often to be overridden by non-constitutional considerations?265

Explaining the differences between indirect federalism rules applied in statutory interpretation, civil rights, and other cases as resulting from underenforcement requires positing: (1) several different federalism norms, (2) several different federalism-related operative propositions, or (3) several distinct sets of instrumental reasons justifying different rules.266 The underenforcement account thus suffers on the consilience and simplicity dimensions.

The underenforcement account also falls short on the conservatism criterion. An underenforcement explanation of our patchwork of indirect federalism doctrines requires imputing to the Court a long-term, trans-substantive federalism implementation strategy actualized in the variety of federalism rules we observe in various contexts. If judges do not agree on the content of federalism norms, it is unlikely that they could consistently agree on the complex set of instrumental reasons that would result in this pattern of rules.267 Implementing such a strategy would require a degree of coordination and agreement on instrumental calculations that stands in tension with what we know about consensus building in ideologically heterogeneous institutions, collective action problems, and multi-member courts.268

Finally, historical judicial struggles with crafting direct federalism rules—most of which have been abandoned because they are unworkable—have highlighted the risk of adjudicatory error involved in making

265. Consider another hypothetical: Courts are asked to shift from rational basis review to strict scrutiny for disability discrimination. Such a change might diminish state government discretion in dealing with disabled individuals—a federalism consideration weighing against the change. If there are no other reasons favoring the change, federalism seems weighty enough and perhaps dispositive. But if the shift was supported by evidence that state disability-related discrimination often proceeds from animus and that stated justifications are pretext such that strict scrutiny would provide a better chance of reaching the right outcome on net, courts might be justified in adopting strict scrutiny. Such calculations are a significant part of the story of how equal protection doctrine reached its current form and they demonstrate one set of conditions under which federalism might be outweighed by instrumental reasons.

266. Young, Federalism Doctrine, supra note 8, at 1819 (arguing that answers to complex institutional capacity questions vary from one federalism issue to another). In attempting to be comprehensive, Young is willing to tolerate some inconsistency in the doctrine—that in itself is no problem, but it does stand in tension with claims that the Court has formulated federalism doctrine according to a strategy that explains all the doctrines. Such a strategy may be too much to ascribe to the Court, at least over more than a few years.

267. See supra notes 43–51, 255–262 and accompanying text (noting these disagreements). 268. Professor Young argues, for example, that much federalism doctrine in recent decades

results from “incompletely theorized agreements.” See Young, Federalism Doctrine, supra note 8, at 1761. See generally Cass R. Sunstein, Incompletely Theorized Agreements, 108 HARV. L. REV. 1733 (1995).



federalism doctrine.269 Commentators who have examined federalism-based interpretive canons suggest that formulating such rules is as rife with the potential for adjudicatory error as is the formulation of substantive federalism rules.270 Even assuming away the debate about the existence of our “rough balance” federalism norm, the vagueness of the constitutionally required balance will still plague any attempt to craft implementing rules, direct or indirect. The rough equivalence of instrumental reasons favoring direct and indirect doctrine shows that the underenforcement account may be inadequate to explain the courts’ choice to focus on indirect federalism rules in recent decades. This may be the reason that Professors Eskridge and Frickey—the only scholars I am aware of to systematically address the possibility that federalism-influenced interpretive rules may be examples of judicial underenforcement of federalism norms—reject the underenforcement account as inadequate.271 Their alternative explanation is judicial activism; mine is based on the variable normativity hypothesis.

The influence of interpretive reasons on constitutional doctrine is well accepted and the influence of instrumental reasons is increasingly acknowledged. The underenforcement account is tempting—the objections I have presented are not dispositive. The test of my alternative account will be whether there are reasons to prefer it to the underenforcement account; but I do not argue that it is preferable because the underenforcement account is somehow clearly wrong. Some might view the federalism issues considered in indirect cases as instrumental matters or their functional equivalent;272 that view, if correct, suggests that federalism is treated as defeasible in these cases for the same reason that instrumental considerations are defeasible. But this does not properly capture federalism’s status as a constitutional norm—again, it reproduces the basic question of how a constitutional norm can function defeasibly—and thus cannot fully explain indirect federalism doctrines. Federalism seems, intuitively, like a constitutional matter that is weightier than other instrumental reasons. And the standard model of constitutional adjudication

269. See Garcia v. San Antonio Metro. Transit Auth., 469 U.S. 528, 550 (1985) (highlighting judicial struggles to construct “principled constitutional limitations” based on federalism norms). Again, this same kind of consideration suggests that underenforcement may not fully explain the state of judicial enforcement of other structural constitutional norms. See Eskridge & Frickey, supra note 4, at 633 (noting that “similar [error-rate related] reasoning underlies the Court’s failure to enforce the non-delegation doctrine and the separation of powers”).

270. Eskridge & Frickey, supra note 4, at 633 (arguing that the judicial error concerns potentially justifying “underenforcement of federalism norms through” substantive doctrine “are equally valid arguments for underenforcement of federalism norms through statutory interpretation”).

271. See id. at 632–46 (making a case against the underenforcement explanation, arguing that “even our best theory for the Court’s practice”—the two output thesis and underenforcement—”fails upon further examination”).

272. Cf. Sager, supra note 10, at 1219–22 (listing structural concerns, including federalism, among the instrumental reasons in doctrinal formulation).



has developed only far enough to defend instrumental calculation in doctrine-making. In the next section, I argue that other kinds of reasons also may legitimately influence doctrine and may better capture federalism’s role in indirect cases.

C. Federalism as a “Quasi-Constitutional” Reason

We may answer our central question as follows: DF is true—federalism norms function defeasibly in some circumstances—because federalism norms in some circumstances have less than the superordinate normative status of standard constitutional norms. This explanation best fits the phenomena—it says that indirect federalism rules are just what they appear to be. To defend this claim, we must answer the following questions: (1) Is the idea of a constitutional norm carrying varying normative priority consistent with what we know about the constitutional system? (2) Can we specify the criteria that determine whether federalism norms will function defeasibly? And, (3) are there reasons to think that this account is preferable to others and particularly to the underenforcement account?

1. Intermediate Normative Priority

To find analytic tools with which to describe federalism’s function in indirect cases, we may benefit from theoretical work on other legal norms with non-standard normative status. Some constitutional theorists are now exploring the extent to which norms outside the canonical constitutional document function as parts of our Constitution.273 Three basic functions of a constitution are structuring government institutions, establishing rights, and entrenching those institutions and rights against easy change; but “under our modern institutional arrangements, the first two of these functions are no longer exclusively, or even primarily, performed by constitutional norms.”274 They are now often performed by ordinary statutes, regulations, executive orders, and so forth.275 Government institutions are structured by administrative statutes like the Federal

273. See generally WILLIAM N. ESKRIDGE, JR. & JOHN FEREJOHN, A REPUBLIC OF STATUTES: THE NEW AMERICAN CONSTITUTION (2010) [hereinafter ESKRIDGE & FEREJOHN, REPUBLIC]; William N. Eskridge, Jr., America’s Statutory “constitution”, 41 U.C. DAVIS L. REV. 1 (2007); Eskridge & Ferejohn, Super-Statutes, supra note 14; Gillian E. Metzger, Ordinary Administrative Law as Constitutional Common Law, 110 COLUM. L. REV. 479 (2010); Ernest A. Young, The Constitution Outside the Constitution, 117 YALE L.J. 408 (2007) [hereinafter Young, Constitution Outside].

274. Young, Constitution Outside, supra note 273, at 412. 275. See, e.g., Eskridge & Ferejohn, Super-Statutes, supra note 14, at 1215; Young, Constitution

Outside, supra note 273, at 412–13 (discussing the Federal Communications Act).



Communications Act276 (FCA); and rights are protected by statutes like Titles VII and IX of the Civil Rights Act.277

Three relevant theses emerge in this literature: First, the categories of “constitutional” and “ordinary” law are more fluid than they seem; laws like the FCA278 may effectively switch categories of normative force, at least in one direction—from ordinary to constitutional law. Some commentators go only this far, arguing that the existence of extracanonical materials with constitutional functions should prompt constitutional scholarship to expand its focus.279 Others press on to suggest implications for our conventional hierarchy of legal priority. One difference between the ordinary and constitutional law categories, of course, is that constitutional law norms are “higher” law; the former have greater normative priority than the latter. The existence of extracanonical constitutional norms suggests that normative priority may change if certain conditions are satisfied.280

The possibility of changing normative status leads to the second thesis, which connects extracanonicalism to the literature on informal constitutional change: The conditions under which normative priority may change—e.g., from ordinary to constitutional priority—have to do with the actions, beliefs and practices of legal officials and the public. Professors Eskridge and Ferejohn, for example, argue that a “super statute”—a statute which, over time, takes on something approaching constitutional normative priority—increases in normative priority after “a pattern of statutory

276. See 47 U.S.C. §§ 151 (2006); Young, Constitution Outside, supra note 273, at 412–13. 277. Civil Rights Act of 1964, Pub. L. No. 88-352, 78 Stat. 253 (codified at 42 U.S.C. § 2000e

(2006)). 278. See 47 U.S.C. § 151. 279. See generally Ernest A. Young, The Constitutive and Entrenchment Functions of

Constitutions: A Research Agenda, 10 U. PA. J. CONST. L. 399 (2007). 280. Professor Young wants to “decouple” the constitutive and entrenchment functions of

constitutions, see Young, Constitution Outside, supra note 273, at 413–14, distinguishing him from theorists who “have insisted on treating . . . extracanonical norms as ‘higher law,’ which puts their theories on a collision course with Article V and creates a great deal of pressure to develop an alternative rule of recognition to identify those norms that have achieved this higher status.” Id. at 413–14. Young observes, however, that some extracanonical norms, though not formally entrenched by Article V, are harder to change for pragmatic and political reasons than is ordinary law. See id. at 413 (arguing that “the broad range of important interests, both individual and commercial, that the [Federal Communication] Act balances and protects ensures that it is, as a practical matter, quite difficult to alter in any sort of fundamental way”). Entrenchment in this literature refers to both a norm’s power to trump other norms and the difficulty of changing the norm. Eskridge and Ferejohn’s view is similar—not in rejecting the possibility of ascension to “higher” law status, which they embrace, but in the sense that they view the form of entrenchment that super statutes enjoy as a result of the practical difficulty of fundamental change created by the statutes’ having become axiomatic in public consciousness and legal practice. See Eskridge & Ferejohn, Super-Statutes, supra note 14, at 1229–30. I leave this debate aside here; I am concerned not with the acquisition of constitutional normativity, but with its loss—processes that are connected but not identical.



enactments reflecting a particular normative stance”281 and widespread public acceptance of the statute’s basic purpose and requirements “such that its critics are discredited and its policy and principles become axiomatic for public culture.”282 Professor Ackerman argues that elevating otherwise ordinary norms to constitutional status requires briefer and more dramatic “constitutional moments,” in which the heightened public awareness of and participation in “higher lawmaking” processes may legitimately alter the content of the canonical Constitution.283

To formulate the point more generally: “[T]he success of constitutional law, in both its constitutive and constraining roles, depends on the willingness and ability of powerful social and political actors to make sustainable commitments to abide by and uphold constitutional rules and institutions.”284 The idea that the beliefs and practices of officials and the public determine the content of legal norms and their place in our hierarchy of legal authority is familiar—it is broadly consistent with theories of popular sovereignty and legal positivism.285 Indeed, Eskridge and Ferejohn’s super-statute theory suggests that official acceptance of quasi-constitutional status with public acquiescence is required to move ordinary law up the hierarchy of legal authority—and their contention that the Sherman Act has achieved that heightened normative status suggests that ordinary laws need not perform any constitutive function to move up the normative hierarchy.286 This demonstrates both the potential breadth of the set of extracanonical norms capable of gaining enhanced normative priority

281. Eskridge & Ferejohn, Super-Statutes, supra note 14, at 1227; see id. at 1216 (“A super-statute is a law or series of laws that (1) seeks to establish a new normative or institutional framework for state policy, and (2) over time does ‘stick’ in the public culture such that (3) the super-statute and its institutional or normative principles have a broad effect on the law—including an effect beyond the four corners of the statute. Super-statutes are typically enacted only after lengthy normative debate about a vexing social or economic problem . . . . The law must also provide a robust solution, a standard, or a norm over time, such that its earlier critics are discredited and its policy and principles become axiomatic for the public culture.” (emphasis added)).

282. Super statutes need not even perform constitutive functions; the Sherman Act, for example, is non-constitutive but nevertheless is a super statute because its principle has been entrenched into public culture. See Eskridge & Ferejohn, Super-Statutes, supra note 14, at 1216–19.

283. See generally 1 BRUCE ACKERMAN, WE THE PEOPLE: FOUNDATIONS 6 (1991). 284. Daryl J. Levinson, Parchment and Politics: The Positive Puzzle of Constitutional

Commitment, 124 HARV. L. REV. 657, 662 (2011). Levinson generalizes this point beyond constitutionalism, arguing that all stable legal norms and institutions are stable only because of the acceptance and acquiescence in practice of officials and the public. See id. at 663. On this view, the distinction between ordinary and higher law amounts to no more than a distinction in practice—norms systematically treated like constitutional norms have that status, and those systematically treated as ordinary are ordinary.

285. I think the first two points cohere with common sense and sentiment. On the parallel with legal positivism, see generally HART, supra note 193, at 111 (arguing that a necessary condition for the existence of a legal system is common acceptance by legal officials of the rule of recognition establishing the system’s criteria of legal validity).

286. See Eskridge & Ferejohn, Super-Statutes, supra note 14, at 1231–37.



and the power of legal practice to alter our normative hierarchy. This is distinct from the claim that the law just is whatever courts say that it is in a particular case287—these views suggest that widespread patterns signaling officials’ convergence on norms may legitimize those norms qua norms and determine their status, not that officials’ talk of norms is pretext for exercises of unfettered discretion. And that suggests that judicial treatment of federalism norms is evidence—not necessarily conclusive, but all things equal, evidence nonetheless—in our inquiry about the normative status of federalism norms.288 Departures from axioms of the standard model in refining our descriptive account of constitutional norms and adjudication thus may be justified where long-term official practice shows that the standard account is incomplete, as we have seen with federalism. DF’s apparent truth may have important implications for its truth.289

The third relevant thesis from the extracanonicalist literature is that legal norms may have intermediate normative priority, located somewhere between ordinary and constitutional law on the hierarchy of legal authority. A super statute, for example, may overcome an ordinary statute with which it conflicts but be trumped by a conflicting constitutional requirement.290 Eskridge and Ferejohn call their super statutes “quasi-constitutional law” to highlight their intermediate normative status.291 This thesis blurs the boundary between ordinary and constitutional law and complicates those categories internally. Some ordinary law norms may have sufficient normative force to trump others; those might be considered “extraordinary” or “quasi-constitutional” norms. We might infer, mutatis mutandis, that some canonical constitutional norms may over time become sufficiently limited in trumping power to lose higher-law status.

Consistent with these theses, it seems reasonable to suppose that the normative priority of canonical constitutional norms might descend to the intermediate normative category under some conditions. On the extracanonicalist reasoning, a necessary and sufficient condition should be a robust pattern of official action, with which the public acquiesces,

287. See supra notes 238–243 and accompanying text. 288. Extracanonicalists similarly rely on observations of practice. Eskridge and Ferejohn

formulated the super-statute account in part to explain patterns of unusual deference and broad application of certain statutory provisions. Eskridge & Ferejohn, Super-Statutes, supra note 14, at 1226–27.

289. There is a distinction between what courts think respecting federalism’s normativity and what the truth is. Here, I need only show that courts act as if they believe federalism is variably normative such that it can apply defeasibly in a variety of cases. For those who take official practice to establish the content of legal norms, this seems to be evidence about the truth.

290. Cf. Eskridge & Ferejohn, Super-Statutes, supra note 14, at 1229–30 (emphasizing the durability requirement for super-statute status).

291. Id. at 1216–17 (arguing that “super statutes” are “‘quasi-constitutional’—fundamental and trumping like constitutional law, but more tentative and susceptible to override or alteration by the legislature or determined judges and administrators”).



treating the canonical norm as one of intermediate priority some of the time. While constitutional theory has so far been concerned with the process by which norms ascend the hierarchy from ordinary to intermediate or constitutional priority; the premises on which that dynamic is hypothesized are general. In principle, nothing precludes the conclusion that the same process may reduce a canonical norm to one of intermediate or ordinary normative priority. And the claim that norms have in fact moved upward suggests that downward movement may also have occurred. It takes time to establish a pattern of official conduct sufficient to suggest an accepted deviation from the standard model—an odd decision on an odd day will not do. This requirement seems satisfied in the case of federalism’s defeasibility. Even if we assume that federalism norms were originally determinative in every application, indirect federalism decisions have issued for at least a century, suggesting that our legal officials do, as a matter of durable agreement, accept that federalism norms may be treated as defeasible under certain circumstances. Think, for example, of the rise of the modern administrative state during the New Deal.292 This was perhaps the most significant alteration of the constitutional structure since Reconstruction and it fundamentally altered the balance of federal and state authority.293 Federalism objections to the growth of the administrative state, if voiced early, have receded from all but some corners of the legal academy. Whatever federalism norm violations grounded early objections, officials now clearly view federalism norms as either compatible with the administrative state or overridden by its pragmatic benefits.

To put it slightly differently: If we may conceive of our constitutionalism as incorporating legal materials beyond the canonical document because those extra-canonical materials perform constitutional functions, then perhaps it is also permissible to think that canonical constitutional norms (written or implied by the constitutional text) may in some circumstances perform non-constitutional functions and thus appropriately be considered non-constitutional in those instances. If “trumping” power is a necessary condition for norms to have constitutional status; then federalism in instances where it does not trump all sub-constitutional considerations is, at best, of quasi-constitutional status.294

292. See generally Reuel E. Schiller, Enlarging the Administrative Polity: Administrative Law and the Changing Definition of Pluralism, 1945–1970, 53 VAND. L. REV. 1389, 1393–96 (2000) (canvassing historical literature); Gary Lawson, The Rise and Rise of the Administrative State, 107 HARV. L. REV. 1231 (1994) (giving a critical history, arguing that the modern administrative state is unconstitutional).

293. See, e.g., Benjamin & Young, supra note 187, at 2113–14 (arguing that the New Deal era of agency expansion poses serious problems for our conceptions of the constitutional structure).

294. This is not necessarily the best understanding of the necessary conditions for “constitutional” norms. It is not the understanding that prevails in many common law countries. See Young, supra note 279, at 404–05 (discussing the British and American constitutions).



The term “quasi-constitutional” in Eskridge and Ferejohn’s usage is apt here since federalism often occupies a similar but distinct intermediate place in our system’s normative hierarchy.

One might object that the potential for ordinary law norms to ascend to intermediate or constitutional normative status does not entail that canonical constitutional norms may—in a mirror-image process—descend to ordinary law status. The asymmetry is a product of the fact that Article V appears to protect canonical norms in their canonical status. But accepting the possibility of extracanonical constitutional norms requires accepting that at least “functionally” constitutional norms may be added outside the Article V amendment process.295 Still, one might object that public acceptance of additional fundamental norms is one thing, it might validate them; but public acceptance altering the actual content (or implication) of the canonical constitutional document—which is required to demote federalism norms to occasional quasi-constitutional status—is another. Professor Ackerman and others, however, have long disputed the exclusivity of the Article V process, arguing that real change to the content of the Constitution has been accomplished through somewhat more grandiose instances of the same official and public acceptance mechanisms emphasized by the extracanonicalists.296 At the very least it is not clear that change to canonical norms can never be accomplished through a process of changing official practices and public views—many argue that they can and have repeatedly been changed by that process. If this is a legitimate mechanism of constitutional change, then the indirect federalism rules—their development and refinement—are evidence that federalism norms have been changed by just this process.

Underenforced status differs from the quasi-constitutional status I am ascribing to federalism norms in a conceptual sense: The latter is a characteristic of the norms themselves—federalism in some instances just is not a fully trumping constitutional norm. The former, by contrast, is a characteristic of the doctrine courts develop to implement constitutional norms—federalism in direct cases is a fully trumping norm, but instrumental concerns may motivate courts to adopt rules that invalidate fewer than all violations.297 This might not amount to a practical difference in some cases, but the conceptual distinction and the category of quasi-constitutional norms it identifies have significant implications for constitutional theory.

Before previewing those implications, however, I briefly return to the two remaining core theory-building questions set out above: Can we say

295. See supra notes 279–289 and accompanying text. 296. See supra notes 281–288 and accompanying text. 297. See supra notes 243–253 and accompanying text.



anything about the conditions under which federalism’s normative priority will vary; and are there reasons to prefer the variable normativity account to the underenforcement account?

2. Variable Normative Priority

To account for phenomena like the “canonicalization” or “constitutionalization” of ordinary legal norms and, indeed, the apparent “de-canonicalization” under certain conditions of canonical federalism norms, we need to add categories between “higher” and “ordinary” priority on the standard model’s normative hierarchy. I have designated one level of intermediate normative priority as “quasi-constitutional,” following Eskridge and Ferejohn, and have posited that our system’s basic federalism norm occupies this station in some cases. We still need to know something about the conditions under which federalism norms may vary in normative force. Again, we may derive these by generalizing from observations of judicial practice.

Official and public treatment of beliefs about legal norms over time are important to determining norms’ proper status in the normative hierarchy. Thus we might distinguish levels of normative force according to situation-types in which legal officials engage federalism norms. Theorists suggest that the situation-type model is one important way in which individuals, including of course judges, experience and think about phenomena in the world.298 Decisions involving federalism norms appear to describe a continuum of normative force varying with the situation-type: On one end are situations that present clear and direct threats to important characteristics of the constitutional structure but have relatively low public policy significance or effect on the enforcement of other constitutional norms. In those situations, courts tend to treat federalism norms as decisive reasons for invalidating government action. Think here of the commandeering cases in which the Court emphasizes the severity of the threat to structural stability but in which it was also clear that Congress likely could get obtain equivalent substantive policy outcomes without the structural threat by using the spending power.299 Toward the other end of the spectrum are the situations that federalism compatibilists often hold up

298. The American Legal Realists argued that judges decide cases primarily based on the situation-type presented rather than the existence of binding, rationally determinate legal norms. See KARL LLEWELLYN, THE COMMON LAW TRADITION: DECIDING APPEALS 124–26 (1960), as discussed in Leiter, supra note 193. On situation types in other contexts, see generally JON BARWISE & JOHN PERRY, SITUATIONS AND ATTITUDES (1983). There is something interesting here about the connection between positive legal theory and brain science suggested by the argument that public and official beliefs about norm content may be dispositive, but I leave that point aside for now.

299. See supra notes 52–67 and accompanying text.



as proof of the need for flexibility in the constitutional structure,300 where the systematic threat is small but rigid adherence to conventional structures threatens the effective implementation of distinct constitutional norms—often rights-bearing norms—or of significant public policy goals.301 In the typically indirect federalism cases, the pattern of official conduct over time is to treat federalism as weighty but defeasible, flexible, or at least a non-prohibitory influence on doctrinal development.302 Think here of preemption and conditional spending cases involving significant federal regulatory, foreign relations or national security interests in which federalism appears as an influence on soft, deferential interpretive canons.303 At the end of this side of the spectrum are cases in which federalism sometimes appears as an influence on doctrine that implements a distinct constitutional norm like the Equal Protection or Due Process Clauses, which carry perhaps the most significant federal policy implications.304 There, federalism seems to have its lowest normative force, functioning as just one of many considerations, on par even with non-legal concerns about comparative institutional capacity, interbranch friction or adjudicatory error, that influence doctrinal formulation.305

Federalism norms with varying normativity appear desirable. From a constitutional design perspective, it makes sense to build federal systems with multiple overlapping safeguards.306 Our system is built that way, with the judicial safeguards for federalism as well as safeguards that are political and procedural in nature, the “popular” safeguard that Madison highlighted in arguing that the level of government capable of commanding the allegiance of the polity would be safeguarded against incursions by the

300. See generally Pursley, Federalism Compatibilists, supra note 15 (describing compatibilists’ normative claims).

301. Some commentators argue that the most significant threats to federalism, those that diminish state regulatory autonomy, are now addressed mostly by indirect doctrines—preemption and conditional spending doctrines, for example—while direct federalism doctrines address actions that are either of little importance to systemic stability or that almost never occur. See, e.g., Young, Two Federalisms, supra note 1, at 50–65, 130–60 (comparing the significance of the different actions federalism doctrines are designed to address). The Court rhetoric in the sovereign immunity and commandeering cases, however, suggests that it, at least, believes that it is addressing the most important federalism problems with direct rules and leaving less important matters for indirect rules. See, e.g., Fed. Mar. Comm’n v. S.C. Ports Auth., 535 U.S. 743, 751–52 (2002) (explaining that sovereign immunity is an “integral component of [states’] residuary and inviolable sovereignty”).

302. See Pursley, Federalism Compatibilists, supra note 15, at 1370–75. 303. See supra notes 127–153 and accompanying text; see, e.g., Wyeth v. Levine, 555 U.S. 555

(2009) (applying presumption against preemption in case about preemption under the federal Food, Drug & Cosmetic Act); Crosby v. Nat’l Foreign Trade Council, 530 U.S. 363 (2000) (omitting presumption in foreign relations preemption case).

304. See supra notes 109–132 and accompanying text. 305. See supra notes 106–121 and accompanying text. 306. See Jenna Bednar, The Political Science of Federalism, 7 ANN. REV. L. & SOC. SCI. 269,

279–80 (2011) (“Ideally, the boundaries of federalism are regulated by a system of complimentary safeguards such that each component compensates for the others’ weaknesses.”).



other, and safeguards flowing from the possibility of state “pushback” against federal overreaching in the context of cooperative regulatory programs.307 Each of these safeguards is imperfect—”prone to failure by considering biased information, such as legally circumscribed evidence in the courtroom or mass perceptions that shape the response of political or popular safeguards.”308 A designer would want to ensure that in situations where one type of safeguard is likely to fail—e.g., in cases where complex questions of overlapping regulatory authority are better suited to legislative than judicial judgment such that the judicial safeguard may suffer information failures—other safeguards will function properly. One way to accomplish this is with norms that trigger the various safeguards under different circumstances. Where policymaking questions concerning the allocation of concurrent authority are at issue in circumstances of grave public concern, we should want the political and popular safeguards of federalism to prevent destabilizing encroachment by one government or another; where either shirking or overenforcement is a problem cooperative regulatory enforcement involving multiple levels of government, we should want the state pushback safeguard to be the primary enforcement mechanism for federalism norms; and so forth.

As the Court has acknowledged, judicial enforcement of federalism norms in instances where courts are ill suited to the task can undermine the proper functioning of the other federalism safeguards.309 Perhaps, then, we should hope to see indirect federalism rules applied in a manner that limits judicial intervention where other safeguards are better suited to the task of enforcing federalism norms and the judicial safeguard is weak. Put differently, the idea that our constitutional federalism norms are such that they trigger both decisive judicial application as in direct cases and something less than that—sometimes, perhaps, in the form of judicial deference to other federalism safeguards—is broadly consistent with current views on optimal design features of federal systems and the parallels of the idealized systems described in the constitutional design literature with our own federal system. Increasingly accepted design principles suggest that it is beneficial for federalism norms to have the sort of variable normative force—at least from the judicial perspective—that I have proposed here. Indeed, given the complexity of a system of multiple federalism safeguards and the default expectation that government institutions will seek to guard their own turf rather than defer to coordinate

307. THE FEDERALIST NO. 46, at 322 (James Madison) (J.E. Cooke ed., 1977) (discussing the function of fluctuating citizen loyalty in the system of federalism safeguards); Gerken, supra note 39 (describing the process by which states exercise power in cooperative regulatory relationships).

308. Bednar, supra note 306, at 280. 309. See Garcia v. San Antonio Metro. Transit Auth., 469 U.S. 528, 531 (1985); United States v.

Lopez, 514 U.S. 549, 571–78 (1995) (Kennedy, J., concurring).,



actors, variably normative federalism norms seem to be a promising solution for calibrating the intensity of judicial federalism enforcement to account for the shortcomings of the judicial role and the functioning of other enforcement mechanisms. The alternative of relying on courts to properly determine when and how much to underenforce uniformly decisive federalism norms seems to leave too much to the influence of institutional self-interest.310

Federalism’s strength as a reason, both generally and in the indirect cases, appears to vary with the significance of the threat to structural stability and with the significance of the non-federalism constitutional norms and the substantive policy concerns at stake in the challenged action. The distinction between direct and indirect federalism rules reveals a pattern of judicial invocation of normatively decisive federalism rules where the court perceives that basic systemic stability or character is threatened. Indirect federalism rules appear in situations that present less danger to the system but involve significant negative consequences if rigid federalism rules are enforced.311 The large question that remains, and that I do not purport to answer here, is whether it is legitimate for courts to treat constitutional norms as having variable normative force for these particular reasons. I have sketched an account on which that practice is explicable, internally coherent, and consistent with well-settled views in constitutional theory. No theory of informal constitutional change yet mediates among officials’ inner motivations for accepting different norms and normative propositions—indeed, the legitimacy question here is significant because it is implicated in all theories of informal constitutional change.

Among the possible explanations canvassed here, this variable normativity account of federalism norms provides the simplest answer to the question of DF’s truth—it suggests that things are as they appear and requires neither an ontology of constitutional principles nor a large and complex effort to uncover the instrumental calculations leading courts to shape federalism doctrine as they have. It succeeds at consilience, explaining every instance in which federalism appears to be considered in a judicial opinion as determinative norm, weighty process-oriented consideration, or otherwise as a reason shaping doctrinal development.

310. Political scientists analyzing federal systems have noted that “the complexities of federalism” in the American system and others “require systems-level analysis.” See Bednar, supra note 306, at 280.

311. I have expressed this variation as a continuum because it does not seem whether there is a “lowest ebb” of federalism’s normative force in indirect cases. One might think that, as a clearly constitutional norm, federalism would carry special weight such that it would always outweigh at least instrumental considerations. But it is not clear that this is right; in any event examining federalism norms’ actual comparative normative force at different points along the continuum is a task for another day.



Facts about the structural and policy stakes in each case determine whether federalism is appropriately treated as fully normative or defeasible. And the variable normativity explanation is conservative: It is consistent with widely accepted views about legal systems, developing ideas about extracanonical constitutionalism and informal constitutional change, and our most nuanced theories of federalism’s role in constitutional adjudication. It provides an answer to the analytically prior positive question of federalism’s defeasibility in indirect cases and thereby fills a significant analytic gap. Indeed, the only accounts of federalism that appear to be in tension with this view involve controversial claims about the content of federalism norms and proposals for significant doctrinal change; denial of the legitimacy of considering federalism issues in non-federalism cases; or denial of the causal force of constitutional norms in adjudication altogether.

One might nevertheless wonder whether the variable normativity account is necessary at all. Perhaps our basic federalism norms are simply so vague that courts can define away numerous potential violations as non-interference or insignificant interference with the federal structure. When added to the possibility of underenforcement for instrumental reasons, norm vagueness might explain federalism’s seemingly diminished significance as a reason in indirect federalism cases. While there may well be an explanatory role for vagueness here, this argument does not obviously explain how federalism norms have sufficient trans-substantive influence to effect doctrine in cases not primarily concerned with violations of federalism norms. More importantly, a vague constitutional norm is still a constitutional norm that, on the standard model, must operate decisively when it applies—the vagueness account, therefore, also fails to explain federalism’s defeasibility in indirect cases.

If we accept that federalism norms have influential but non-decisive normative force in nearly every case, we can examine and construct an account of the reasons courts have for examining federalism questions in non-federalism cases. The variable normativity account both diminishes the force of federalism norms in many cases and suggests that it may be safely considered in a wide variety of contexts because courts need not worry that taking up the federalism issue will mandate a decision contrary to that most consistent with the norms primarily at issue. The constitutional structure is almost always somehow implicated in constitutional cases. Even a vague norm will have some clear applications.312 If federalism norms are treated

312. See Young, Federalism Doctrine, supra note 8, at 1775–83 (noting the vagueness of basic federalism norms); Benjamin & Young, supra note 187, at 2117 (“The constitutional structure builds in a vast amount of play in the joints, but we do think attention to certain hardwired institutions and



as bearing superordinate normative force in every instance, then wherever federalism norms are raised and impose clear requirements, there is no escaping them without sacrificing the appearance of legitimacy. That dynamic runs counter to our observations in indirect federalism cases, where even clear federalism requirements—such as the requirement of congressional involvement in decisions to preempt state law—may give way to sufficiently strong non-federalism considerations.313

Finally, consider a normative point about the legitimacy of quasi-constitutional norms: Entrenching structural norms with variable normative priority is a good idea from a constitutional design perspective if, as in our system, an important design goal is to guarantee the durability of the constitutional structure.314 Our history demonstrates that circumstances can change quickly and dramatically; for any constitutional system to endure in its basic character through periods of social, political or economic upheaval, it must be able to bend without breaking—fostering durability requires mechanisms by which structures may adapt to changing circumstances. There is fairly broad consensus today that Article V’s amendment process is too onerous to provide for sufficient adaptability.315 One way to ensure flexibility is to enact just this sort of variably normative federalism norm, effectively building capacity for adaptation into the system from the beginning. Making it variable in normative priority seems a better approach than simply making norms vague and hoping that future generations of officials will discern the need for adaptation and underenforce or otherwise soften through interpretation rigid structures as needed to accommodate various “crises of human affairs.”316 While many have argued that the Constitution’s vagueness on structure creates the needed dynamism; one problem with vague norms is that they effectively delegate to future officials the authority to determine norm-content in all applications. The risk is that adaptation may be undertaken without necessity or with disregard for the structure’s basic character even if that character was meant to endure. A more concrete norm can guide and constrain future decision-makers, with greater force on ordinary questions

processes is non-optional. Prominent among them is the centrality of Congress’s legislative decisions in our constitutional scheme.” (emphasis added)).

313. Commentators have noted the seemingly arbitrary manner in which the presumption against preemption is applied in some preemption cases but not in others. See, e.g., Davis, supra note 5, at 1013); Young, Two Federalisms, supra note 1, at 130–34.

314. See Pursley, Dormancy, supra note 20, at 512–14 (arguing that durability as a basic intended feature of the constitutional system); Eskridge & Ferejohn, Super-Statutes, supra note 14, at 1268 (“In order for any constitution to be enduring, it must be dynamic). See generally sourced cited supra note 273.

315. See, e.g., SANFORD LEVINSON, OUR UNDEMOCRATIC CONSTITUTION: WHERE THE

CONSTITUTION GOES WRONG (AND HOW THE PEOPLE CAN CORRECT IT) (2006). 316. McCulloch v. Maryland, 17 U.S. (4 Wheat.) 316, 415 (1819).



on which adaptation is unnecessary or undesirable. Variable normative priority allows for flexibility while also providing some constraint and concreteness.

This is more than a good reason for thinking that the variable normativity hypothesis supports the better account of federalism norms’ roles in adjudication; it also adds to the positive account of how we commit durably to constitutional norms. Those commitments may be softened in their tendency to give rise to the inter-temporal difficulty (or “dead hand” problem)—a famous doubt about the legitimacy of ascribing power over people today to members of prior generations by hewing flexibility to antiquated structures.317 Norms of variable normativite priority render those past commitments flexible enough, perhaps, to dispel this worry.

CONCLUSION

Compatibilist federalism arguments have become increasingly prevalent.318 Scholars accept that constitutional structures must flex to adapt to changed circumstances and work up suggestions for how federalism doctrine might be adapted to improve outcomes in a variety of contexts. What that literature lacks is an account of why it is legitimate to assume that federalism doctrine can “flex.” The possibility of instrumentally motivated underenforcement of federalism norms provides a partial explanation, but, as we have observed, underenforcement does not obviously explain some instances of non-standard judicial treatment of federalism norms. And underenforcement’s explanatory capacity seems insufficient to justify every compatabilist proposal. The idea that federalism norms function quasi-constitutionally under certain circumstances provides another possible conceptual foundation for the compatibilists’ normative project. It also adds a salient example of non-standard constitutional norms to the burgeoning literature on the scope and content of our constitutional canon. The benefits of hypothesizing variable normative priority for understanding federalism doctrine—explaining a great deal of otherwise seemingly ad hoc doctrine making—contributes to the literature on the nature of constitutional doctrine and the practice of constitutional doctrine-making a new category of reasons courts may legitimately account for in formulating doctrinal rules. There is much yet to explore, including how thinking of federalism as a quasi-constitutional norm may advance theoretical controversies about the substance of federalism itself.

317. See Levinson, supra note 284 (discussing commitment puzzles). 318. See generally Pursley, Federalism Compatibilists, supra note 15.

3 pursley 801 - 865 as of 6-4 63... · 3 pursley 801 - 865 as of 6-4 (do not delete) 6/4/2012 3:36...

Documents